From: Muchun Song <hidden> Date: 2021-02-25 13:24:24
Hi all,
This patch series will free some vmemmap pages(struct page structures)
associated with each hugetlbpage when preallocated to save memory.
In order to reduce the difficulty of the first version of code review.
From this version, we disable PMD/huge page mapping of vmemmap if this
feature was enabled. This accutualy eliminate a bunch of the complex code
doing page table manipulation. When this patch series is solid, we cam add
the code of vmemmap page table manipulation in the future.
The struct page structures (page structs) are used to describe a physical
page frame. By default, there is a one-to-one mapping from a page frame to
it's corresponding page struct.
The HugeTLB pages consist of multiple base page size pages and is supported
by many architectures. See hugetlbpage.rst in the Documentation directory
for more details. On the x86 architecture, HugeTLB pages of size 2MB and 1GB
are currently supported. Since the base page size on x86 is 4KB, a 2MB
HugeTLB page consists of 512 base pages and a 1GB HugeTLB page consists of
4096 base pages. For each base page, there is a corresponding page struct.
Within the HugeTLB subsystem, only the first 4 page structs are used to
contain unique information about a HugeTLB page. HUGETLB_CGROUP_MIN_ORDER
provides this upper limit. The only 'useful' information in the remaining
page structs is the compound_head field, and this field is the same for all
tail pages.
By removing redundant page structs for HugeTLB pages, memory can returned to
the buddy allocator for other uses.
When the system boot up, every 2M HugeTLB has 512 struct page structs which
size is 8 pages(sizeof(struct page) * 512 / PAGE_SIZE).
HugeTLB struct pages(8 pages) page frame(8 pages)
+-----------+ ---virt_to_page---> +-----------+ mapping to +-----------+
| | | 0 | -------------> | 0 |
| | +-----------+ +-----------+
| | | 1 | -------------> | 1 |
| | +-----------+ +-----------+
| | | 2 | -------------> | 2 |
| | +-----------+ +-----------+
| | | 3 | -------------> | 3 |
| | +-----------+ +-----------+
| | | 4 | -------------> | 4 |
| 2MB | +-----------+ +-----------+
| | | 5 | -------------> | 5 |
| | +-----------+ +-----------+
| | | 6 | -------------> | 6 |
| | +-----------+ +-----------+
| | | 7 | -------------> | 7 |
| | +-----------+ +-----------+
| |
| |
| |
+-----------+
The value of page->compound_head is the same for all tail pages. The first
page of page structs (page 0) associated with the HugeTLB page contains the 4
page structs necessary to describe the HugeTLB. The only use of the remaining
pages of page structs (page 1 to page 7) is to point to page->compound_head.
Therefore, we can remap pages 2 to 7 to page 1. Only 2 pages of page structs
will be used for each HugeTLB page. This will allow us to free the remaining
6 pages to the buddy allocator.
Here is how things look after remapping.
HugeTLB struct pages(8 pages) page frame(8 pages)
+-----------+ ---virt_to_page---> +-----------+ mapping to +-----------+
| | | 0 | -------------> | 0 |
| | +-----------+ +-----------+
| | | 1 | -------------> | 1 |
| | +-----------+ +-----------+
| | | 2 | ----------------^ ^ ^ ^ ^ ^
| | +-----------+ | | | | |
| | | 3 | ------------------+ | | | |
| | +-----------+ | | | |
| | | 4 | --------------------+ | | |
| 2MB | +-----------+ | | |
| | | 5 | ----------------------+ | |
| | +-----------+ | |
| | | 6 | ------------------------+ |
| | +-----------+ |
| | | 7 | --------------------------+
| | +-----------+
| |
| |
| |
+-----------+
When a HugeTLB is freed to the buddy system, we should allocate 6 pages for
vmemmap pages and restore the previous mapping relationship.
Apart from 2MB HugeTLB page, we also have 1GB HugeTLB page. It is similar
to the 2MB HugeTLB page. We also can use this approach to free the vmemmap
pages.
In this case, for the 1GB HugeTLB page, we can save 4094 pages. This is a
very substantial gain. On our server, run some SPDK/QEMU applications which
will use 1024GB hugetlbpage. With this feature enabled, we can save ~16GB
(1G hugepage)/~12GB (2MB hugepage) memory.
Because there are vmemmap page tables reconstruction on the freeing/allocating
path, it increases some overhead. Here are some overhead analysis.
1) Allocating 10240 2MB hugetlb pages.
a) With this patch series applied:
# time echo 10240 > /proc/sys/vm/nr_hugepages
real 0m0.166s
user 0m0.000s
sys 0m0.166s
# bpftrace -e 'kprobe:alloc_fresh_huge_page { @start[tid] = nsecs; }
kretprobe:alloc_fresh_huge_page /@start[tid]/ { @latency = hist(nsecs -
@start[tid]); delete(@start[tid]); }'
Attaching 2 probes...
@latency:
[8K, 16K) 5476 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@|
[16K, 32K) 4760 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ |
[32K, 64K) 4 | |
b) Without this patch series:
# time echo 10240 > /proc/sys/vm/nr_hugepages
real 0m0.067s
user 0m0.000s
sys 0m0.067s
# bpftrace -e 'kprobe:alloc_fresh_huge_page { @start[tid] = nsecs; }
kretprobe:alloc_fresh_huge_page /@start[tid]/ { @latency = hist(nsecs -
@start[tid]); delete(@start[tid]); }'
Attaching 2 probes...
@latency:
[4K, 8K) 10147 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@|
[8K, 16K) 93 | |
Summarize: this feature is about ~2x slower than before.
2) Freeing 10240 2MB hugetlb pages.
a) With this patch series applied:
# time echo 0 > /proc/sys/vm/nr_hugepages
real 0m0.213s
user 0m0.000s
sys 0m0.213s
# bpftrace -e 'kprobe:free_pool_huge_page { @start[tid] = nsecs; }
kretprobe:free_pool_huge_page /@start[tid]/ { @latency = hist(nsecs -
@start[tid]); delete(@start[tid]); }'
Attaching 2 probes...
@latency:
[8K, 16K) 6 | |
[16K, 32K) 10227 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@|
[32K, 64K) 7 | |
b) Without this patch series:
# time echo 0 > /proc/sys/vm/nr_hugepages
real 0m0.081s
user 0m0.000s
sys 0m0.081s
# bpftrace -e 'kprobe:free_pool_huge_page { @start[tid] = nsecs; }
kretprobe:free_pool_huge_page /@start[tid]/ { @latency = hist(nsecs -
@start[tid]); delete(@start[tid]); }'
Attaching 2 probes...
@latency:
[4K, 8K) 6805 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@|
[8K, 16K) 3427 |@@@@@@@@@@@@@@@@@@@@@@@@@@ |
[16K, 32K) 8 | |
Summarize: The overhead of __free_hugepage is about ~2-3x slower than before.
Although the overhead has increased, the overhead is not significant. Like Mike
said, "However, remember that the majority of use cases create hugetlb pages at
or shortly after boot time and add them to the pool. So, additional overhead is
at pool creation time. There is no change to 'normal run time' operations of
getting a page from or returning a page to the pool (think page fault/unmap)".
Despite the overhead and in addition to the memory gains from this series. The
following data is obtained by Joao Martins. Very thanks to his effort.
There's an additional benefit which is page (un)pinners will see an improvement
and Joao presumes because there are fewer memmap pages and thus the tail/head
pages are staying in cache more often.
Out of the box Joao saw (when comparing linux-next against linux-next + this series)
with gup_test and pinning a 16G hugetlb file (with 1G pages):
get_user_pages(): ~32k -> ~9k
unpin_user_pages(): ~75k -> ~70k
Usually any tight loop fetching compound_head(), or reading tail pages data (e.g.
compound_head) benefit a lot. There's some unpinning inefficiencies Joao was
fixing[0], but with that in added it shows even more:
unpin_user_pages(): ~27k -> ~3.8k
[0] https://lore.kernel.org/linux-mm/20210204202500.26474-1-joao.m.martins@oracle.com/
Todo:
- Free all of the tail vmemmap pages
Now for the 2MB HugrTLB page, we only free 6 vmemmap pages. we really can
free 7 vmemmap pages. In this case, we can see 8 of the 512 struct page
structures has beed set PG_head flag. If we can adjust compound_head()
slightly and make compound_head() return the real head struct page when
the parameter is the tail struct page but with PG_head flag set.
In order to make the code evolution route clearer. This feature can can be
a separate patch after this patchset is solid.
- Support for other architectures (e.g. aarch64).
- Enable PMD/huge page mapping of vmemmap even if this feature was enabled.
Changelog in v16 -> v17:
- Fix issues suggested by Mike and Oscar.
- Update commit log suggested by Michal.
Thanks to Mike, David H and Michal's suggestions and review.
Changelog in v15 -> v16:
- Use GFP_KERNEL to allocate vmemmap pages.
Thanks to Mike, David H and Michal's suggestions.
Changelog in v14 -> v15:
- Fix some issues suggested by Oscar. Thanks to Oscar.
- Add numbers which Joao Martins tested to cover letter. Thanks to his effort.
Changelog in v13 -> v14:
- Refuse to free the HugeTLB page when the system is under memory pressure.
- Use GFP_ATOMIC to allocate vmemmap pages instead of GFP_KERNEL.
- Rebase to linux-next 20210202.
- Fix and add some comments for vmemmap_remap_free().
Thanks to Oscar, Mike, David H and David R's suggestions and review.
Changelog in v12 -> v13:
- Remove VM_WARN_ON_PAGE macro.
- Add more comments in vmemmap_pte_range() and vmemmap_remap_free().
Thanks to Oscar and Mike's suggestions and review.
Changelog in v11 -> v12:
- Move VM_WARN_ON_PAGE to a separate patch.
- Call __free_hugepage() with hugetlb_lock (See patch #5.) to serialize
with dissolve_free_huge_page(). It is to prepare for patch #9.
- Introduce PageHugeInflight. See patch #9.
Changelog in v10 -> v11:
- Fix compiler error when !CONFIG_HUGETLB_PAGE_FREE_VMEMMAP.
- Rework some comments and commit changes.
- Rework vmemmap_remap_free() to 3 parameters.
Thanks to Oscar and Mike's suggestions and review.
Changelog in v9 -> v10:
- Fix a bug in patch #11. Thanks to Oscar for pointing that out.
- Rework some commit log or comments. Thanks Mike and Oscar for the suggestions.
- Drop VMEMMAP_TAIL_PAGE_REUSE in the patch #3.
Thank you very much Mike and Oscar for reviewing the code.
Changelog in v8 -> v9:
- Rework some code. Very thanks to Oscar.
- Put all the non-hugetlb vmemmap functions under sparsemem-vmemmap.c.
Changelog in v7 -> v8:
- Adjust the order of patches.
Very thanks to David and Oscar. Your suggestions are very valuable.
Changelog in v6 -> v7:
- Rebase to linux-next 20201130
- Do not use basepage mapping for vmemmap when this feature is disabled.
- Rework some patchs.
[PATCH v6 08/16] mm/hugetlb: Free the vmemmap pages associated with each hugetlb page
[PATCH v6 10/16] mm/hugetlb: Allocate the vmemmap pages associated with each hugetlb page
Thanks to Oscar and Barry.
Changelog in v5 -> v6:
- Disable PMD/huge page mapping of vmemmap if this feature was enabled.
- Simplify the first version code.
Changelog in v4 -> v5:
- Rework somme comments and code in the [PATCH v4 04/21] and [PATCH v4 05/21].
Thanks to Mike and Oscar's suggestions.
Changelog in v3 -> v4:
- Move all the vmemmap functions to hugetlb_vmemmap.c.
- Make the CONFIG_HUGETLB_PAGE_FREE_VMEMMAP default to y, if we want to
disable this feature, we should disable it by a boot/kernel command line.
- Remove vmemmap_pgtable_{init, deposit, withdraw}() helper functions.
- Initialize page table lock for vmemmap through core_initcall mechanism.
Thanks for Mike and Oscar's suggestions.
Changelog in v2 -> v3:
- Rename some helps function name. Thanks Mike.
- Rework some code. Thanks Mike and Oscar.
- Remap the tail vmemmap page with PAGE_KERNEL_RO instead of PAGE_KERNEL.
Thanks Matthew.
- Add some overhead analysis in the cover letter.
- Use vmemap pmd table lock instead of a hugetlb specific global lock.
Changelog in v1 -> v2:
- Fix do not call dissolve_compound_page in alloc_huge_page_vmemmap().
- Fix some typo and code style problems.
- Remove unused handle_vmemmap_fault().
- Merge some commits to one commit suggested by Mike.
Muchun Song (9):
mm: memory_hotplug: factor out bootmem core functions to
bootmem_info.c
mm: hugetlb: introduce a new config HUGETLB_PAGE_FREE_VMEMMAP
mm: hugetlb: free the vmemmap pages associated with each HugeTLB page
mm: hugetlb: alloc the vmemmap pages associated with each HugeTLB page
mm: hugetlb: set the PageHWPoison to the raw error page
mm: hugetlb: add a kernel parameter hugetlb_free_vmemmap
mm: hugetlb: introduce nr_free_vmemmap_pages in the struct hstate
mm: hugetlb: gather discrete indexes of tail page
mm: hugetlb: optimize the code with the help of the compiler
Documentation/admin-guide/kernel-parameters.txt | 14 ++
Documentation/admin-guide/mm/hugetlbpage.rst | 11 +
arch/x86/mm/init_64.c | 13 +-
fs/Kconfig | 6 +
include/linux/bootmem_info.h | 65 ++++++
include/linux/hugetlb.h | 47 +++-
include/linux/hugetlb_cgroup.h | 19 +-
include/linux/memory_hotplug.h | 27 ---
include/linux/mm.h | 5 +
mm/Makefile | 2 +
mm/bootmem_info.c | 124 ++++++++++
mm/hugetlb.c | 176 +++++++++++---
mm/hugetlb_vmemmap.c | 293 ++++++++++++++++++++++++
mm/hugetlb_vmemmap.h | 51 +++++
mm/memory_hotplug.c | 116 ----------
mm/sparse-vmemmap.c | 280 ++++++++++++++++++++++
mm/sparse.c | 1 +
17 files changed, 1062 insertions(+), 188 deletions(-)
create mode 100644 include/linux/bootmem_info.h
create mode 100644 mm/bootmem_info.c
create mode 100644 mm/hugetlb_vmemmap.c
create mode 100644 mm/hugetlb_vmemmap.h
--
2.11.0
From: Muchun Song <hidden> Date: 2021-02-25 13:25:14
Move bootmem info registration common API to individual bootmem_info.c.
And we will use {get,put}_page_bootmem() to initialize the page for the
vmemmap pages or free the vmemmap pages to buddy in the later patch.
So move them out of CONFIG_MEMORY_HOTPLUG_SPARSE. This is just code
movement without any functional change.
Signed-off-by: Muchun Song <redacted>
Acked-by: Mike Kravetz <redacted>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Reviewed-by: David Hildenbrand <redacted>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
---
arch/x86/mm/init_64.c | 3 +-
include/linux/bootmem_info.h | 40 +++++++++++++
include/linux/memory_hotplug.h | 27 ---------
mm/Makefile | 1 +
mm/bootmem_info.c | 124 +++++++++++++++++++++++++++++++++++++++++
mm/memory_hotplug.c | 116 --------------------------------------
mm/sparse.c | 1 +
7 files changed, 168 insertions(+), 144 deletions(-)
create mode 100644 include/linux/bootmem_info.h
create mode 100644 mm/bootmem_info.c
@@ -1571,7 +1572,7 @@ int __meminit vmemmap_populate(unsigned long start, unsigned long end, int node,returnerr;}-#if defined(CONFIG_MEMORY_HOTPLUG_SPARSE) && defined(CONFIG_HAVE_BOOTMEM_INFO_NODE)+#ifdef CONFIG_HAVE_BOOTMEM_INFO_NODEvoidregister_page_bootmem_memmap(unsignedlongsection_nr,structpage*start_page,unsignedlongnr_pages){
@@ -18,18 +18,6 @@ struct vmem_altmap;#ifdef CONFIG_MEMORY_HOTPLUGstructpage*pfn_to_online_page(unsignedlongpfn);-/*-*Typesforfreebootmemstoredinpage->lru.next.Thesehavetobein-*somerandomrangeinunsignedlongspacefordebuggingpurposes.-*/-enum{-MEMORY_HOTPLUG_MIN_BOOTMEM_TYPE=12,-SECTION_INFO=MEMORY_HOTPLUG_MIN_BOOTMEM_TYPE,-MIX_SECTION_INFO,-NODE_INFO,-MEMORY_HOTPLUG_MAX_BOOTMEM_TYPE=NODE_INFO,-};-/* Types for control the zone type of onlined and offlined memory */enum{/* Offline the memory. */
From: Muchun Song <hidden> Date: 2021-02-25 13:25:17
The option HUGETLB_PAGE_FREE_VMEMMAP allows for the freeing of
some vmemmap pages associated with pre-allocated HugeTLB pages.
For example, on X86_64 6 vmemmap pages of size 4KB each can be
saved for each 2MB HugeTLB page. 4094 vmemmap pages of size 4KB
each can be saved for each 1GB HugeTLB page.
When a HugeTLB page is allocated or freed, the vmemmap array
representing the range associated with the page will need to be
remapped. When a page is allocated, vmemmap pages are freed
after remapping. When a page is freed, previously discarded
vmemmap pages must be allocated before remapping.
The config option is introduced early so that supporting code
can be written to depend on the option. The initial version of
the code only provides support for x86-64.
Like other code which frees vmemmap, this config option depends on
HAVE_BOOTMEM_INFO_NODE. The routine register_page_bootmem_info() is
used to register bootmem info. Therefore, make sure
register_page_bootmem_info is enabled if HUGETLB_PAGE_FREE_VMEMMAP
is defined.
Signed-off-by: Muchun Song <redacted>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Acked-by: Mike Kravetz <redacted>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
---
arch/x86/mm/init_64.c | 2 +-
fs/Kconfig | 6 ++++++
2 files changed, 7 insertions(+), 1 deletion(-)
From: Muchun Song <hidden> Date: 2021-02-25 13:25:17
Every HugeTLB has more than one struct page structure. We __know__ that
we only use the first 4(HUGETLB_CGROUP_MIN_ORDER) struct page structures
to store metadata associated with each HugeTLB.
There are a lot of struct page structures associated with each HugeTLB
page. For tail pages, the value of compound_head is the same. So we can
reuse first page of tail page structures. We map the virtual addresses
of the remaining pages of tail page structures to the first tail page
struct, and then free these page frames. Therefore, we need to reserve
two pages as vmemmap areas.
When we allocate a HugeTLB page from the buddy, we can free some vmemmap
pages associated with each HugeTLB page. It is more appropriate to do it
in the prep_new_huge_page().
The free_vmemmap_pages_per_hpage(), which indicates how many vmemmap
pages associated with a HugeTLB page can be freed, returns zero for
now, which means the feature is disabled. We will enable it once all
the infrastructure is there.
Signed-off-by: Muchun Song <redacted>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
---
include/linux/bootmem_info.h | 27 +++++-
include/linux/mm.h | 3 +
mm/Makefile | 1 +
mm/hugetlb.c | 3 +
mm/hugetlb_vmemmap.c | 219 +++++++++++++++++++++++++++++++++++++++++++
mm/hugetlb_vmemmap.h | 20 ++++
mm/sparse-vmemmap.c | 207 ++++++++++++++++++++++++++++++++++++++++
7 files changed, 479 insertions(+), 1 deletion(-)
create mode 100644 mm/hugetlb_vmemmap.c
create mode 100644 mm/hugetlb_vmemmap.h
@@ -27,8 +27,215 @@#include<linux/spinlock.h>#include<linux/vmalloc.h>#include<linux/sched.h>+#include<linux/pgtable.h>+#include<linux/bootmem_info.h>+#include<asm/dma.h>#include<asm/pgalloc.h>+#include<asm/tlbflush.h>++/**+*vmemmap_remap_walk-walkvmemmappagetable+*+*@remap_pte:calledforeachlowest-levelentry(PTE).+*@reuse_page:thepagewhichisreusedforthetailvmemmappages.+*@reuse_addr:thevirtualaddressofthe@reuse_pagepage.+*@vmemmap_pages:thelistheadofthevmemmappagesthatcanbefreed.+*/+structvmemmap_remap_walk{+void(*remap_pte)(pte_t*pte,unsignedlongaddr,+structvmemmap_remap_walk*walk);+structpage*reuse_page;+unsignedlongreuse_addr;+structlist_head*vmemmap_pages;+};++staticvoidvmemmap_pte_range(pmd_t*pmd,unsignedlongaddr,+unsignedlongend,+structvmemmap_remap_walk*walk)+{+pte_t*pte;++pte=pte_offset_kernel(pmd,addr);++/*+*Thereuse_pageisfound'first'intablewalkbeforewestart+*remapping(whichiscalling@walk->remap_pte).+*/+if(!walk->reuse_page){+BUG_ON(pte_none(*pte));+BUG_ON(walk->reuse_addr!=addr);++walk->reuse_page=pte_page(*pte++);+/*+*Becausethereuseaddressispartoftherangethatweare+*walking,skipthereuseaddressrange.+*/+addr+=PAGE_SIZE;+}++for(;addr!=end;addr+=PAGE_SIZE,pte++){+BUG_ON(pte_none(*pte));++walk->remap_pte(pte,addr,walk);+}+}++staticvoidvmemmap_pmd_range(pud_t*pud,unsignedlongaddr,+unsignedlongend,+structvmemmap_remap_walk*walk)+{+pmd_t*pmd;+unsignedlongnext;++pmd=pmd_offset(pud,addr);+do{+BUG_ON(pmd_none(*pmd)||pmd_leaf(*pmd));++next=pmd_addr_end(addr,end);+vmemmap_pte_range(pmd,addr,next,walk);+}while(pmd++,addr=next,addr!=end);+}++staticvoidvmemmap_pud_range(p4d_t*p4d,unsignedlongaddr,+unsignedlongend,+structvmemmap_remap_walk*walk)+{+pud_t*pud;+unsignedlongnext;++pud=pud_offset(p4d,addr);+do{+BUG_ON(pud_none(*pud));++next=pud_addr_end(addr,end);+vmemmap_pmd_range(pud,addr,next,walk);+}while(pud++,addr=next,addr!=end);+}++staticvoidvmemmap_p4d_range(pgd_t*pgd,unsignedlongaddr,+unsignedlongend,+structvmemmap_remap_walk*walk)+{+p4d_t*p4d;+unsignedlongnext;++p4d=p4d_offset(pgd,addr);+do{+BUG_ON(p4d_none(*p4d));++next=p4d_addr_end(addr,end);+vmemmap_pud_range(p4d,addr,next,walk);+}while(p4d++,addr=next,addr!=end);+}++staticvoidvmemmap_remap_range(unsignedlongstart,unsignedlongend,+structvmemmap_remap_walk*walk)+{+unsignedlongaddr=start;+unsignedlongnext;+pgd_t*pgd;++VM_BUG_ON(!IS_ALIGNED(start,PAGE_SIZE));+VM_BUG_ON(!IS_ALIGNED(end,PAGE_SIZE));++pgd=pgd_offset_k(addr);+do{+BUG_ON(pgd_none(*pgd));++next=pgd_addr_end(addr,end);+vmemmap_p4d_range(pgd,addr,next,walk);+}while(pgd++,addr=next,addr!=end);++/*+*Weonlychangethemappingofthevmemmapvirtualaddressrange+*[@start+PAGE_SIZE,end),soweonlyneedtoflushtheTLBwhich+*belongstotherange.+*/+flush_tlb_kernel_range(start+PAGE_SIZE,end);+}++/*+*Freeavmemmappage.Avmemmappagecanbeallocatedfromthememblock+*allocatororbuddyallocator.IfthePG_reservedflagisset,itmeans+*thatitallocatedfromthememblockallocator,justfreeitviathe+*free_bootmem_page().Otherwise,use__free_page().+*/+staticinlinevoidfree_vmemmap_page(structpage*page)+{+if(PageReserved(page))+free_bootmem_page(page);+else+__free_page(page);+}++/* Free a list of the vmemmap pages */+staticvoidfree_vmemmap_page_list(structlist_head*list)+{+structpage*page,*next;++list_for_each_entry_safe(page,next,list,lru){+list_del(&page->lru);+free_vmemmap_page(page);+}+}++staticvoidvmemmap_remap_pte(pte_t*pte,unsignedlongaddr,+structvmemmap_remap_walk*walk)+{+/*+*Remapthetailpagesasread-onlytocatchillegalwriteoperation+*tothetailpages.+*/+pgprot_tpgprot=PAGE_KERNEL_RO;+pte_tentry=mk_pte(walk->reuse_page,pgprot);+structpage*page=pte_page(*pte);++list_add(&page->lru,walk->vmemmap_pages);+set_pte_at(&init_mm,addr,pte,entry);+}++/**+*vmemmap_remap_free-remapthevmemmapvirtualaddressrange[@start,@end)+*tothepagewhich@reuseismappedto,thenfreevmemmap+*whichtherangearemappedto.+*@start:startaddressofthevmemmapvirtualaddressrangethatwewant+*toremap.+*@end:endaddressofthevmemmapvirtualaddressrangethatwewantto+*remap.+*@reuse:reuseaddress.+*+*Note:Thisfunctiondependsonvmemmapbeingbasepagemapped.Pleasemake+*surethatwedisablePMDmappingofvmemmappageswhencallingthisfunction.+*/+voidvmemmap_remap_free(unsignedlongstart,unsignedlongend,+unsignedlongreuse)+{+LIST_HEAD(vmemmap_pages);+structvmemmap_remap_walkwalk={+.remap_pte=vmemmap_remap_pte,+.reuse_addr=reuse,+.vmemmap_pages=&vmemmap_pages,+};++/*+*Inordertomakeremappingroutinemostefficientforthehugepages,+*theroutineofvmemmappagetablewalkinghasthefollowingrules+*(seemoredetailsfromthevmemmap_pte_range()):+*+*-Therange[@start,@end)andtherange[@reuse,@reuse+PAGE_SIZE)+*shouldbecontinuous.+*-The@reuseaddressispartoftherange[@reuse,@end)thatweare+*walkingwhichispassedtovmemmap_remap_range().+*-The@reuseaddressisthefirstinthecompleterange.+*+*Soweneedtomakesurethat@startand@reusemeettheaboverules.+*/+BUG_ON(start-reuse!=PAGE_SIZE);++vmemmap_remap_range(reuse,end,&walk);+free_vmemmap_page_list(&vmemmap_pages);+}/**Allocateablockofmemorytobeusedtobackthevirtualmemorymap
From: Muchun Song <hidden> Date: 2021-02-25 13:27:00
When we free a HugeTLB page to the buddy allocator, we should allocate
the vmemmap pages associated with it. But we may cannot allocate vmemmap
pages when the system is under memory pressure, in this case, we just
refuse to free the HugeTLB page instead of looping forever trying to
allocate the pages. This changes some behavior (list below) on some
corner cases.
1) Failing to free a huge page triggered by the user (decrease nr_pages).
Need try again later by the user.
2) Failing to free a surplus huge page when freed by the application.
Try again later when freeing a huge page next time.
3) Failing to dissolve a free huge page on ZONE_MOVABLE via
offline_pages().
This is a bit unfortunate if we have plenty of ZONE_MOVABLE memory
but are low on kernel memory. For example, migration of huge pages
would still work, however, dissolving the free page does not work.
This is a corner cases. When the system is that much under memory
pressure, offlining/unplug can be expected to fail. This is
unfortunate because it prevents from the memory offlining which
shouldn't happen for movable zones. People depending on the memory
hotplug and movable zone should carefuly consider whether savings
on unmovable memory are worth losing their hotplug functionality
in some situations.
4) Failing to dissolve a huge page on CMA/ZONE_MOVABLE via
alloc_contig_range() - once we have that handling in place. Mainly
affects CMA and virtio-mem.
Similar to 3). virito-mem will handle migration errors gracefully.
CMA might be able to fallback on other free areas within the CMA
region.
Vmemmap pages are allocated from the page freeing context. In order for
those allocations to be not disruptive (e.g. trigger oom killer)
__GFP_NORETRY is used. hugetlb_lock is dropped for the allocation
because a non sleeping allocation would be too fragile and it could fail
too easily under memory pressure. GFP_ATOMIC or other modes to access
memory reserves is not used because we want to prevent consuming
reserves under heavy hugetlb freeing.
Signed-off-by: Muchun Song <redacted>
---
Documentation/admin-guide/mm/hugetlbpage.rst | 8 +++
include/linux/mm.h | 2 +
mm/hugetlb.c | 92 +++++++++++++++++++++-------
mm/hugetlb_vmemmap.c | 32 ++++++----
mm/hugetlb_vmemmap.h | 23 +++++++
mm/sparse-vmemmap.c | 75 ++++++++++++++++++++++-
6 files changed, 197 insertions(+), 35 deletions(-)
@@ -60,6 +60,10 @@ HugePages_Surp the pool above the value in ``/proc/sys/vm/nr_hugepages``. The maximum number of surplus huge pages is controlled by``/proc/sys/vm/nr_overcommit_hugepages``.+ Note: When the feature of freeing unused vmemmap pages associated+ with each hugetlb page is enabled, the number of surplus huge pages+ may be temporarily larger than the maximum number of surplus huge+ pages when the system is under memory pressure. Hugepagesize is the default hugepage size (in Kb). Hugetlb
@@ -80,6 +84,10 @@ returned to the huge page pool when freed by a task. A user with root privileges can dynamically allocate more or free some persistent huge pages by increasing or decreasing the value of ``nr_hugepages``.+Note: When the feature of freeing unused vmemmap pages associated with each+hugetlb page is enabled, we can fail to free the huge pages triggered by+the user when ths system is under memory pressure. Please try again later.+ Pages that are used as huge pages are reserved inside the kernel and cannot be used for other purposes. Huge pages cannot be swapped out under memory pressure.
@@ -1404,9 +1443,9 @@ static void __free_huge_page(struct page *page)}elseif(h->surplus_huge_pages_node[nid]){/* remove the page from active list */list_del(&page->lru);-update_and_free_page(h,page);h->surplus_huge_pages--;h->surplus_huge_pages_node[nid]--;+update_and_free_page(h,page);}else{arch_clear_hugepage_flags(page);enqueue_huge_page(h,page);
@@ -237,6 +238,78 @@ void vmemmap_remap_free(unsigned long start, unsigned long end,free_vmemmap_page_list(&vmemmap_pages);}+staticvoidvmemmap_restore_pte(pte_t*pte,unsignedlongaddr,+structvmemmap_remap_walk*walk)+{+pgprot_tpgprot=PAGE_KERNEL;+structpage*page;+void*to;++BUG_ON(pte_page(*pte)!=walk->reuse_page);++page=list_first_entry(walk->vmemmap_pages,structpage,lru);+list_del(&page->lru);+to=page_to_virt(page);+copy_page(to,(void*)walk->reuse_addr);++set_pte_at(&init_mm,addr,pte,mk_pte(page,pgprot));+}++staticintalloc_vmemmap_page_list(unsignedlongstart,unsignedlongend,+gfp_tgfp_mask,structlist_head*list)+{+unsignedlongnr_pages=(end-start)>>PAGE_SHIFT;+intnid=page_to_nid((structpage*)start);+structpage*page,*next;++while(nr_pages--){+page=alloc_pages_node(nid,gfp_mask,0);+if(!page)+gotoout;+list_add_tail(&page->lru,list);+}++return0;+out:+list_for_each_entry_safe(page,next,list,lru)+__free_pages(page,0);+return-ENOMEM;+}++/**+*vmemmap_remap_alloc-remapthevmemmapvirtualaddressrange[@start,end)+*tothepagewhichisfromthe@vmemmap_pages+*respectively.+*@start:startaddressofthevmemmapvirtualaddressrangethatwewant+*toremap.+*@end:endaddressofthevmemmapvirtualaddressrangethatwewantto+*remap.+*@reuse:reuseaddress.+*@gpf_mask:GFPflagforallocatingvmemmappages.+*/+intvmemmap_remap_alloc(unsignedlongstart,unsignedlongend,+unsignedlongreuse,gfp_tgfp_mask)+{+LIST_HEAD(vmemmap_pages);+structvmemmap_remap_walkwalk={+.remap_pte=vmemmap_restore_pte,+.reuse_addr=reuse,+.vmemmap_pages=&vmemmap_pages,+};++/* See the comment in the vmemmap_remap_free(). */+BUG_ON(start-reuse!=PAGE_SIZE);++might_sleep_if(gfpflags_allow_blocking(gfp_mask));++if(alloc_vmemmap_page_list(start,end,gfp_mask,&vmemmap_pages))+return-ENOMEM;++vmemmap_remap_range(reuse,end,&walk);++return0;+}+/**Allocateablockofmemorytobeusedtobackthevirtualmemorymap*ortobackthepagetablesthatareusedtocreatethemapping.
From: Muchun Song <hidden> Date: 2021-02-25 13:27:12
Because we reuse the first tail vmemmap page frame and remap it
with read-only, we cannot set the PageHWPosion on some tail pages.
So we can use the head[4].private (There are at least 128 struct
page structures associated with the optimized HugeTLB page, so
using head[4].private is safe) to record the real error page index
and set the raw error page PageHWPoison later.
Signed-off-by: Muchun Song <redacted>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Acked-by: David Rientjes <rientjes@google.com>
---
mm/hugetlb.c | 80 ++++++++++++++++++++++++++++++++++++++++++++++++++++++------
1 file changed, 72 insertions(+), 8 deletions(-)
From: Muchun Song <hidden> Date: 2021-02-25 13:27:15
Add a kernel parameter hugetlb_free_vmemmap to enable the feature of
freeing unused vmemmap pages associated with each hugetlb page on boot.
We disables PMD mapping of vmemmap pages for x86-64 arch when this
feature is enabled. Because vmemmap_remap_free() depends on vmemmap
being base page mapped.
Signed-off-by: Muchun Song <redacted>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Reviewed-by: Barry Song <redacted>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
---
Documentation/admin-guide/kernel-parameters.txt | 14 ++++++++++++++
Documentation/admin-guide/mm/hugetlbpage.rst | 3 +++
arch/x86/mm/init_64.c | 8 ++++++--
include/linux/hugetlb.h | 19 +++++++++++++++++++
mm/hugetlb_vmemmap.c | 24 ++++++++++++++++++++++++
5 files changed, 66 insertions(+), 2 deletions(-)
@@ -1557,6 +1557,20 @@ Documentation/admin-guide/mm/hugetlbpage.rst. Format: size[KMG]+ hugetlb_free_vmemmap=+ [KNL] When CONFIG_HUGETLB_PAGE_FREE_VMEMMAP is set,+ this controls freeing unused vmemmap pages associated+ with each HugeTLB page. When this option is enabled,+ we disable PMD/huge page mapping of vmemmap pages which+ increase page table pages. So if a user/sysadmin only+ uses a small number of HugeTLB pages (as a percentage+ of system memory), they could end up using more memory+ with hugetlb_free_vmemmap on as opposed to off.+ Format: { on | off (default) }++ on: enable the feature+ off: disable the feature+ hung_task_panic= [KNL] Should the hung task detector generate panics. Format: 0 | 1
@@ -153,6 +153,9 @@ default_hugepagesz will all result in 256 2M huge pages being allocated. Valid default huge page size is architecture dependent.+hugetlb_free_vmemmap+ When CONFIG_HUGETLB_PAGE_FREE_VMEMMAP is set, this enables freeing+ unused vmemmap pages associated with each HugeTLB page. When multiple huge page sizes are supported, ``/proc/sys/vm/nr_hugepages`` indicates the current number of pre-allocated huge pages of the default size.
@@ -1557,7 +1558,8 @@ int __meminit vmemmap_populate(unsigned long start, unsigned long end, int node,{interr;-if(end-start<PAGES_PER_SECTION*sizeof(structpage))+if((is_hugetlb_free_vmemmap_enabled()&&!altmap)||+end-start<PAGES_PER_SECTION*sizeof(structpage))err=vmemmap_populate_basepages(start,end,node,NULL);elseif(boot_cpu_has(X86_FEATURE_PSE))err=vmemmap_populate_hugepages(start,end,node,altmap);
@@ -1585,6 +1587,8 @@ void register_page_bootmem_memmap(unsigned long section_nr,pmd_t*pmd;unsignedintnr_pmd_pages;structpage*page;+boolbase_mapping=!boot_cpu_has(X86_FEATURE_PSE)||+is_hugetlb_free_vmemmap_enabled();for(;addr<end;addr=next){pte_t*pte=NULL;
@@ -1610,7 +1614,7 @@ void register_page_bootmem_memmap(unsigned long section_nr,}get_page_bootmem(section_nr,pud_page(*pud),MIX_SECTION_INFO);-if(!boot_cpu_has(X86_FEATURE_PSE)){+if(base_mapping){next=(addr+PAGE_SIZE)&PAGE_MASK;pmd=pmd_offset(pud,addr);if(pmd_none(*pmd))
From: Muchun Song <hidden> Date: 2021-02-25 13:27:27
All the infrastructure is ready, so we introduce nr_free_vmemmap_pages
field in the hstate to indicate how many vmemmap pages associated with
a HugeTLB page that can be freed to buddy allocator. And initialize it
in the hugetlb_vmemmap_init(). This patch is actual enablement of the
feature.
Signed-off-by: Muchun Song <redacted>
Acked-by: Mike Kravetz <redacted>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
---
include/linux/hugetlb.h | 3 +++
mm/hugetlb.c | 1 +
mm/hugetlb_vmemmap.c | 25 +++++++++++++++++++++++++
mm/hugetlb_vmemmap.h | 10 ++++++----
4 files changed, 35 insertions(+), 4 deletions(-)
From: Muchun Song <hidden> Date: 2021-02-25 13:27:46
For HugeTLB page, there are more metadata to save in the struct page.
But the head struct page cannot meet our needs, so we have to abuse
other tail struct page to store the metadata. In order to avoid
conflicts caused by subsequent use of more tail struct pages, we can
gather these discrete indexes of tail struct page. In this case, it
will be easier to add a new tail page index later.
There are only (RESERVE_VMEMMAP_SIZE / sizeof(struct page)) struct
page structs that can be used when CONFIG_HUGETLB_PAGE_FREE_VMEMMAP,
so add a BUILD_BUG_ON to catch invalid usage of the tail struct page.
Signed-off-by: Muchun Song <redacted>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
---
include/linux/hugetlb.h | 24 ++++++++++++++++++++++--
include/linux/hugetlb_cgroup.h | 19 +++++++++++--------
mm/hugetlb.c | 6 +++---
mm/hugetlb_vmemmap.c | 8 ++++++++
4 files changed, 44 insertions(+), 13 deletions(-)
From: Muchun Song <hidden> Date: 2021-02-25 13:28:29
When the "struct page size" crosses page boundaries we cannot
make use of this feature. Let free_vmemmap_pages_per_hpage()
return zero if that is the case, most of the functions can be
optimized away.
Signed-off-by: Muchun Song <redacted>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
---
include/linux/hugetlb.h | 3 ++-
mm/hugetlb_vmemmap.c | 7 +++++++
mm/hugetlb_vmemmap.h | 6 ++++++
3 files changed, 15 insertions(+), 1 deletion(-)
From: Muchun Song <hidden> Date: 2021-03-01 05:30:55
On Thu, Feb 25, 2021 at 9:24 PM Muchun Song [off-list ref] wrote:
When we free a HugeTLB page to the buddy allocator, we should allocate
the vmemmap pages associated with it. But we may cannot allocate vmemmap
pages when the system is under memory pressure, in this case, we just
refuse to free the HugeTLB page instead of looping forever trying to
allocate the pages. This changes some behavior (list below) on some
corner cases.
1) Failing to free a huge page triggered by the user (decrease nr_pages).
Need try again later by the user.
2) Failing to free a surplus huge page when freed by the application.
Try again later when freeing a huge page next time.
3) Failing to dissolve a free huge page on ZONE_MOVABLE via
offline_pages().
This is a bit unfortunate if we have plenty of ZONE_MOVABLE memory
but are low on kernel memory. For example, migration of huge pages
would still work, however, dissolving the free page does not work.
This is a corner cases. When the system is that much under memory
pressure, offlining/unplug can be expected to fail. This is
unfortunate because it prevents from the memory offlining which
shouldn't happen for movable zones. People depending on the memory
hotplug and movable zone should carefuly consider whether savings
on unmovable memory are worth losing their hotplug functionality
in some situations.
4) Failing to dissolve a huge page on CMA/ZONE_MOVABLE via
alloc_contig_range() - once we have that handling in place. Mainly
affects CMA and virtio-mem.
Similar to 3). virito-mem will handle migration errors gracefully.
CMA might be able to fallback on other free areas within the CMA
region.
Vmemmap pages are allocated from the page freeing context. In order for
those allocations to be not disruptive (e.g. trigger oom killer)
__GFP_NORETRY is used. hugetlb_lock is dropped for the allocation
because a non sleeping allocation would be too fragile and it could fail
too easily under memory pressure. GFP_ATOMIC or other modes to access
memory reserves is not used because we want to prevent consuming
reserves under heavy hugetlb freeing.
Hi,
Since this patch is the only patch that has no reviewed-by tag.
I hope someone (e.g. Mike, Oscar, David or Michal) could review
this. Thanks a lot.
@@ -60,6 +60,10 @@ HugePages_Surp the pool above the value in ``/proc/sys/vm/nr_hugepages``. The maximum number of surplus huge pages is controlled by``/proc/sys/vm/nr_overcommit_hugepages``.+ Note: When the feature of freeing unused vmemmap pages associated+ with each hugetlb page is enabled, the number of surplus huge pages+ may be temporarily larger than the maximum number of surplus huge+ pages when the system is under memory pressure. Hugepagesize is the default hugepage size (in Kb). Hugetlb
@@ -80,6 +84,10 @@ returned to the huge page pool when freed by a task. A user with root privileges can dynamically allocate more or free some persistent huge pages by increasing or decreasing the value of ``nr_hugepages``.+Note: When the feature of freeing unused vmemmap pages associated with each+hugetlb page is enabled, we can fail to free the huge pages triggered by+the user when ths system is under memory pressure. Please try again later.+ Pages that are used as huge pages are reserved inside the kernel and cannot be used for other purposes. Huge pages cannot be swapped out under memory pressure.
@@ -1404,9 +1443,9 @@ static void __free_huge_page(struct page *page)}elseif(h->surplus_huge_pages_node[nid]){/* remove the page from active list */list_del(&page->lru);-update_and_free_page(h,page);h->surplus_huge_pages--;h->surplus_huge_pages_node[nid]--;+update_and_free_page(h,page);}else{arch_clear_hugepage_flags(page);enqueue_huge_page(h,page);
@@ -237,6 +238,78 @@ void vmemmap_remap_free(unsigned long start, unsigned long end,free_vmemmap_page_list(&vmemmap_pages);}+staticvoidvmemmap_restore_pte(pte_t*pte,unsignedlongaddr,+structvmemmap_remap_walk*walk)+{+pgprot_tpgprot=PAGE_KERNEL;+structpage*page;+void*to;++BUG_ON(pte_page(*pte)!=walk->reuse_page);++page=list_first_entry(walk->vmemmap_pages,structpage,lru);+list_del(&page->lru);+to=page_to_virt(page);+copy_page(to,(void*)walk->reuse_addr);++set_pte_at(&init_mm,addr,pte,mk_pte(page,pgprot));+}++staticintalloc_vmemmap_page_list(unsignedlongstart,unsignedlongend,+gfp_tgfp_mask,structlist_head*list)+{+unsignedlongnr_pages=(end-start)>>PAGE_SHIFT;+intnid=page_to_nid((structpage*)start);+structpage*page,*next;++while(nr_pages--){+page=alloc_pages_node(nid,gfp_mask,0);+if(!page)+gotoout;+list_add_tail(&page->lru,list);+}++return0;+out:+list_for_each_entry_safe(page,next,list,lru)+__free_pages(page,0);+return-ENOMEM;+}++/**+*vmemmap_remap_alloc-remapthevmemmapvirtualaddressrange[@start,end)+*tothepagewhichisfromthe@vmemmap_pages+*respectively.+*@start:startaddressofthevmemmapvirtualaddressrangethatwewant+*toremap.+*@end:endaddressofthevmemmapvirtualaddressrangethatwewantto+*remap.+*@reuse:reuseaddress.+*@gpf_mask:GFPflagforallocatingvmemmappages.+*/+intvmemmap_remap_alloc(unsignedlongstart,unsignedlongend,+unsignedlongreuse,gfp_tgfp_mask)+{+LIST_HEAD(vmemmap_pages);+structvmemmap_remap_walkwalk={+.remap_pte=vmemmap_restore_pte,+.reuse_addr=reuse,+.vmemmap_pages=&vmemmap_pages,+};++/* See the comment in the vmemmap_remap_free(). */+BUG_ON(start-reuse!=PAGE_SIZE);++might_sleep_if(gfpflags_allow_blocking(gfp_mask));++if(alloc_vmemmap_page_list(start,end,gfp_mask,&vmemmap_pages))+return-ENOMEM;++vmemmap_remap_range(reuse,end,&walk);++return0;+}+/**Allocateablockofmemorytobeusedtobackthevirtualmemorymap*ortobackthepagetablesthatareusedtocreatethemapping.--
Hi all,
This patch series will free some vmemmap pages(struct page structures)
associated with each hugetlbpage when preallocated to save memory.
In order to reduce the difficulty of the first version of code review.
quoted
From this version, we disable PMD/huge page mapping of vmemmap if this
feature was enabled. This accutualy eliminate a bunch of the complex code
doing page table manipulation. When this patch series is solid, we cam add
the code of vmemmap page table manipulation in the future.
The struct page structures (page structs) are used to describe a physical
page frame. By default, there is a one-to-one mapping from a page frame to
it's corresponding page struct.
The HugeTLB pages consist of multiple base page size pages and is supported
by many architectures. See hugetlbpage.rst in the Documentation directory
for more details. On the x86 architecture, HugeTLB pages of size 2MB and 1GB
are currently supported. Since the base page size on x86 is 4KB, a 2MB
HugeTLB page consists of 512 base pages and a 1GB HugeTLB page consists of
4096 base pages. For each base page, there is a corresponding page struct.
Within the HugeTLB subsystem, only the first 4 page structs are used to
contain unique information about a HugeTLB page. HUGETLB_CGROUP_MIN_ORDER
provides this upper limit. The only 'useful' information in the remaining
page structs is the compound_head field, and this field is the same for all
tail pages.
By removing redundant page structs for HugeTLB pages, memory can returned to
the buddy allocator for other uses.
When the system boot up, every 2M HugeTLB has 512 struct page structs which
size is 8 pages(sizeof(struct page) * 512 / PAGE_SIZE).
HugeTLB struct pages(8 pages) page frame(8 pages)
+-----------+ ---virt_to_page---> +-----------+ mapping to +-----------+
| | | 0 | -------------> | 0 |
| | +-----------+ +-----------+
| | | 1 | -------------> | 1 |
| | +-----------+ +-----------+
| | | 2 | -------------> | 2 |
| | +-----------+ +-----------+
| | | 3 | -------------> | 3 |
| | +-----------+ +-----------+
| | | 4 | -------------> | 4 |
| 2MB | +-----------+ +-----------+
| | | 5 | -------------> | 5 |
| | +-----------+ +-----------+
| | | 6 | -------------> | 6 |
| | +-----------+ +-----------+
| | | 7 | -------------> | 7 |
| | +-----------+ +-----------+
| |
| |
| |
+-----------+
The value of page->compound_head is the same for all tail pages. The first
page of page structs (page 0) associated with the HugeTLB page contains the 4
page structs necessary to describe the HugeTLB. The only use of the remaining
pages of page structs (page 1 to page 7) is to point to page->compound_head.
Therefore, we can remap pages 2 to 7 to page 1. Only 2 pages of page structs
will be used for each HugeTLB page. This will allow us to free the remaining
6 pages to the buddy allocator.
Here is how things look after remapping.
HugeTLB struct pages(8 pages) page frame(8 pages)
+-----------+ ---virt_to_page---> +-----------+ mapping to +-----------+
| | | 0 | -------------> | 0 |
| | +-----------+ +-----------+
| | | 1 | -------------> | 1 |
| | +-----------+ +-----------+
| | | 2 | ----------------^ ^ ^ ^ ^ ^
| | +-----------+ | | | | |
| | | 3 | ------------------+ | | | |
| | +-----------+ | | | |
| | | 4 | --------------------+ | | |
| 2MB | +-----------+ | | |
| | | 5 | ----------------------+ | |
| | +-----------+ | |
| | | 6 | ------------------------+ |
| | +-----------+ |
| | | 7 | --------------------------+
| | +-----------+
| |
| |
| |
+-----------+
When a HugeTLB is freed to the buddy system, we should allocate 6 pages for
vmemmap pages and restore the previous mapping relationship.
Apart from 2MB HugeTLB page, we also have 1GB HugeTLB page. It is similar
to the 2MB HugeTLB page. We also can use this approach to free the vmemmap
pages.
In this case, for the 1GB HugeTLB page, we can save 4094 pages. This is a
very substantial gain. On our server, run some SPDK/QEMU applications which
will use 1024GB hugetlbpage. With this feature enabled, we can save ~16GB
(1G hugepage)/~12GB (2MB hugepage) memory.
Because there are vmemmap page tables reconstruction on the freeing/allocating
path, it increases some overhead. Here are some overhead analysis.
1) Allocating 10240 2MB hugetlb pages.
a) With this patch series applied:
# time echo 10240 > /proc/sys/vm/nr_hugepages
real 0m0.166s
user 0m0.000s
sys 0m0.166s
# bpftrace -e 'kprobe:alloc_fresh_huge_page { @start[tid] = nsecs; }
kretprobe:alloc_fresh_huge_page /@start[tid]/ { @latency = hist(nsecs -
@start[tid]); delete(@start[tid]); }'
Attaching 2 probes...
@latency:
[8K, 16K) 5476 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@|
[16K, 32K) 4760 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ |
[32K, 64K) 4 | |
b) Without this patch series:
# time echo 10240 > /proc/sys/vm/nr_hugepages
real 0m0.067s
user 0m0.000s
sys 0m0.067s
# bpftrace -e 'kprobe:alloc_fresh_huge_page { @start[tid] = nsecs; }
kretprobe:alloc_fresh_huge_page /@start[tid]/ { @latency = hist(nsecs -
@start[tid]); delete(@start[tid]); }'
Attaching 2 probes...
@latency:
[4K, 8K) 10147 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@|
[8K, 16K) 93 | |
Summarize: this feature is about ~2x slower than before.
2) Freeing 10240 2MB hugetlb pages.
a) With this patch series applied:
# time echo 0 > /proc/sys/vm/nr_hugepages
real 0m0.213s
user 0m0.000s
sys 0m0.213s
# bpftrace -e 'kprobe:free_pool_huge_page { @start[tid] = nsecs; }
kretprobe:free_pool_huge_page /@start[tid]/ { @latency = hist(nsecs -
@start[tid]); delete(@start[tid]); }'
Attaching 2 probes...
@latency:
[8K, 16K) 6 | |
[16K, 32K) 10227 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@|
[32K, 64K) 7 | |
b) Without this patch series:
# time echo 0 > /proc/sys/vm/nr_hugepages
real 0m0.081s
user 0m0.000s
sys 0m0.081s
# bpftrace -e 'kprobe:free_pool_huge_page { @start[tid] = nsecs; }
kretprobe:free_pool_huge_page /@start[tid]/ { @latency = hist(nsecs -
@start[tid]); delete(@start[tid]); }'
Attaching 2 probes...
@latency:
[4K, 8K) 6805 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@|
[8K, 16K) 3427 |@@@@@@@@@@@@@@@@@@@@@@@@@@ |
[16K, 32K) 8 | |
Summarize: The overhead of __free_hugepage is about ~2-3x slower than before.
Although the overhead has increased, the overhead is not significant. Like Mike
said, "However, remember that the majority of use cases create hugetlb pages at
or shortly after boot time and add them to the pool. So, additional overhead is
at pool creation time. There is no change to 'normal run time' operations of
getting a page from or returning a page to the pool (think page fault/unmap)".
Despite the overhead and in addition to the memory gains from this series. The
following data is obtained by Joao Martins. Very thanks to his effort.
There's an additional benefit which is page (un)pinners will see an improvement
and Joao presumes because there are fewer memmap pages and thus the tail/head
pages are staying in cache more often.
Out of the box Joao saw (when comparing linux-next against linux-next + this series)
with gup_test and pinning a 16G hugetlb file (with 1G pages):
get_user_pages(): ~32k -> ~9k
unpin_user_pages(): ~75k -> ~70k
Usually any tight loop fetching compound_head(), or reading tail pages data (e.g.
compound_head) benefit a lot. There's some unpinning inefficiencies Joao was
fixing[0], but with that in added it shows even more:
unpin_user_pages(): ~27k -> ~3.8k
[0] https://lore.kernel.org/linux-mm/20210204202500.26474-1-joao.m.martins@oracle.com/
Todo:
- Free all of the tail vmemmap pages
Now for the 2MB HugrTLB page, we only free 6 vmemmap pages. we really can
free 7 vmemmap pages. In this case, we can see 8 of the 512 struct page
structures has beed set PG_head flag. If we can adjust compound_head()
slightly and make compound_head() return the real head struct page when
the parameter is the tail struct page but with PG_head flag set.
In order to make the code evolution route clearer. This feature can can be
a separate patch after this patchset is solid.
- Support for other architectures (e.g. aarch64).
- Enable PMD/huge page mapping of vmemmap even if this feature was enabled.
Tested-by: Chen Huang <redacted>
We are interested in this patch and have tested the patch for x86. Also we made a simple
modification in arm64 and tested for the patch.
1. In x86, we set the total memory of 70G, and use 32G for hugepages then got the result:
------------------------------------------------------------------------------------------------
2M page | 1G page |
----------------------|------------------------|----------------------|------------------------|
enable | disable | enable | disable |
----------------------|------------------------|----------------------|------------------------|
total | used | free | total | used | free |total | used | free | total | used | free |
70855 | 33069 | 37786| 70473 | 33068 | 37405 |70983 | 33068 | 37914| 70473 | 33068 | 37405 |
------------------------------------------------------------------------------------------------
The result is that for 2M hugepage, we can save 382M memory which is correspoinding to the expected
384M memory. For 1G hugepage, we can save 510M memory which is correspoinding to the expected 512M
memory.
2. In arm64, the hack modification is shown below[1]. We set the total memory of 40G, and use 10G
for hugepages then got the result:
------------------------------------------------------------------------------------------------
2M page | 1G page |
----------------------|------------------------|----------------------|------------------------|
enable | disable | enable | disable |
----------------------|------------------------|----------------------|------------------------|
total | used | free | total | used | free |total | used | free | total | used | free |
39,739 | 10279 |29,460| 39579 | 10278 | 29,301|39,699 | 10279 |29,420| 39579 | 10278 | 29,301|
------------------------------------------------------------------------------------------------
The result is that for 2M hugepage, we can save 119M memory which is correspoinding to the expected
120M memory. For 1G hugepage, we can save 159M memory which is correspoinding to the expected 160M
memory.
3. Also we found that when we free and realloc the 1G hugepages, as the vmemmap need realloc pages before
freeing the hugepages, it will decrese the chance to get hugepages. Because the freeed hugepages has
returned to the buddy system and may be used for vmemmap pages.
We think failing to alloc hugepages is normal so this is fine.
[1]: support arm64
From: Mike Kravetz <hidden> Date: 2021-03-03 05:38:40
On 2/25/21 5:21 AM, Muchun Song wrote:
When we free a HugeTLB page to the buddy allocator, we should allocate
the vmemmap pages associated with it. But we may cannot allocate vmemmap
pages when the system is under memory pressure, in this case, we just
refuse to free the HugeTLB page instead of looping forever trying to
allocate the pages. This changes some behavior (list below) on some
cassociated with it.orner cases.
Suggest rewording this as:
When we free a HugeTLB page to the buddy allocator, we need to allocate
the vmemmap pages associated with it. However, we may not be able to
allocate the vmemmap pages when the system is under memory pressure. In
this case, we just refuse to free the HugeTLB page. This changes behavior
in some corner cases as listed below:
1) Failing to free a huge page triggered by the user (decrease nr_pages).
Need try again later by the user.
User needs to try again later.
2) Failing to free a surplus huge page when freed by the application.
Try again later when freeing a huge page next time.
3) Failing to dissolve a free huge page on ZONE_MOVABLE via
offline_pages().
This is a bit unfortunate if we have plenty of ZONE_MOVABLE memory
but are low on kernel memory. For example, migration of huge pages
would still work, however, dissolving the free page does not work.
This is a corner cases. When the system is that much under memory
pressure, offlining/unplug can be expected to fail. This is
unfortunate because it prevents from the memory offlining which
shouldn't happen for movable zones. People depending on the memory
hotplug and movable zone should carefuly consider whether savings
on unmovable memory are worth losing their hotplug functionality
in some situations.
Possible wording change:
This can happen when we have plenty of ZONE_MOVABLE memory, but
not enough kernel memory to allocate vmemmmap pages. We may even
be able to migrate huge page contents, but will not be able to
dissolve the source huge page. This will prevent an offline
operation and is unfortunate as memory offlining is expected to
succeed on movable zones. Users that depend on memory hotplug
to succeed for movable zones should carefully consider whether the
memory savings gained from this feature are worth the risk of
possibly not being able to offline memory in certain situations.
quoted hunk
4) Failing to dissolve a huge page on CMA/ZONE_MOVABLE via
alloc_contig_range() - once we have that handling in place. Mainly
affects CMA and virtio-mem.
Similar to 3). virito-mem will handle migration errors gracefully.
CMA might be able to fallback on other free areas within the CMA
region.
Vmemmap pages are allocated from the page freeing context. In order for
those allocations to be not disruptive (e.g. trigger oom killer)
__GFP_NORETRY is used. hugetlb_lock is dropped for the allocation
because a non sleeping allocation would be too fragile and it could fail
too easily under memory pressure. GFP_ATOMIC or other modes to access
memory reserves is not used because we want to prevent consuming
reserves under heavy hugetlb freeing.
Signed-off-by: Muchun Song <redacted>
---
Documentation/admin-guide/mm/hugetlbpage.rst | 8 +++
include/linux/mm.h | 2 +
mm/hugetlb.c | 92 +++++++++++++++++++++-------
mm/hugetlb_vmemmap.c | 32 ++++++----
mm/hugetlb_vmemmap.h | 23 +++++++
mm/sparse-vmemmap.c | 75 ++++++++++++++++++++++-
6 files changed, 197 insertions(+), 35 deletions(-)
@@ -60,6 +60,10 @@ HugePages_Surp the pool above the value in ``/proc/sys/vm/nr_hugepages``. The maximum number of surplus huge pages is controlled by``/proc/sys/vm/nr_overcommit_hugepages``.+ Note: When the feature of freeing unused vmemmap pages associated+ with each hugetlb page is enabled, the number of surplus huge pages+ may be temporarily larger than the maximum number of surplus huge+ pages when the system is under memory pressure. Hugepagesize is the default hugepage size (in Kb). Hugetlb
@@ -80,6 +84,10 @@ returned to the huge page pool when freed by a task. A user with root privileges can dynamically allocate more or free some persistent huge pages by increasing or decreasing the value of ``nr_hugepages``.+Note: When the feature of freeing unused vmemmap pages associated with each+hugetlb page is enabled, we can fail to free the huge pages triggered by+the user when ths system is under memory pressure. Please try again later.+ Pages that are used as huge pages are reserved inside the kernel and cannot be used for other purposes. Huge pages cannot be swapped out under memory pressure.
@@ -1404,9 +1443,9 @@ static void __free_huge_page(struct page *page) } else if (h->surplus_huge_pages_node[nid]) { /* remove the page from active list */ list_del(&page->lru);- update_and_free_page(h, page); h->surplus_huge_pages--; h->surplus_huge_pages_node[nid]--;+ update_and_free_page(h, page); } else { arch_clear_hugepage_flags(page); enqueue_huge_page(h, page);
@@ -1447,7 +1486,7 @@ void free_huge_page(struct page *page) /* * Defer freeing if in non-task context to avoid hugetlb_lock deadlock. */- if (!in_task()) {+ if (!in_atomic()) {
That should be "if (in_atomic()) instead of "if (!in_atomic())"
Do note that there is an ongoing discussion about calling free_huge_page
in various contexts.
https://lore.kernel.org/linux-mm/000000000000f1c03b05bc43aadc@google.com/
This discussion/issue is independent of this patch. Since that issue
deals with existing code, we will need to come up with a solution there
first. A solution there may impact how free_huge_page is structured and
may impact this patch.
The in_atomic() check is insufficient to handle all cases. It is better
than !in_task(), but still does not cover all cases.
The rest of the patch looks good to me.
--
Mike Kravetz
quoted hunk
/*
* Only call schedule_work() if hpage_freelist is previously
* empty. Otherwise, schedule_work() had been called but the
@@ -1699,8 +1738,7 @@ static int free_pool_huge_page(struct hstate *h, nodemask_t *nodes_allowed, h->surplus_huge_pages--; h->surplus_huge_pages_node[node]--; }- update_and_free_page(h, page);- ret = 1;+ ret = !update_and_free_page(h, page); break; } }
@@ -1713,10 +1751,14 @@ static int free_pool_huge_page(struct hstate *h, nodemask_t *nodes_allowed, * nothing for in-use hugepages and non-hugepages. * This function returns values like below: *- * -EBUSY: failed to dissolved free hugepages or the hugepage is in-use- * (allocated or reserved.)- * 0: successfully dissolved free hugepages or the page is not a- * hugepage (considered as already dissolved)+ * -ENOMEM: failed to allocate vmemmap pages to free the freed hugepages+ * when the system is under memory pressure and the feature of+ * freeing unused vmemmap pages associated with each hugetlb page+ * is enabled.+ * -EBUSY: failed to dissolved free hugepages or the hugepage is in-use+ * (allocated or reserved.)+ * 0: successfully dissolved free hugepages or the page is not a+ * hugepage (considered as already dissolved) */ int dissolve_free_huge_page(struct page *page) {
@@ -237,6 +238,78 @@ void vmemmap_remap_free(unsigned long start, unsigned long end,free_vmemmap_page_list(&vmemmap_pages);}+staticvoidvmemmap_restore_pte(pte_t*pte,unsignedlongaddr,+structvmemmap_remap_walk*walk)+{+pgprot_tpgprot=PAGE_KERNEL;+structpage*page;+void*to;++BUG_ON(pte_page(*pte)!=walk->reuse_page);++page=list_first_entry(walk->vmemmap_pages,structpage,lru);+list_del(&page->lru);+to=page_to_virt(page);+copy_page(to,(void*)walk->reuse_addr);++set_pte_at(&init_mm,addr,pte,mk_pte(page,pgprot));+}++staticintalloc_vmemmap_page_list(unsignedlongstart,unsignedlongend,+gfp_tgfp_mask,structlist_head*list)+{+unsignedlongnr_pages=(end-start)>>PAGE_SHIFT;+intnid=page_to_nid((structpage*)start);+structpage*page,*next;++while(nr_pages--){+page=alloc_pages_node(nid,gfp_mask,0);+if(!page)+gotoout;+list_add_tail(&page->lru,list);+}++return0;+out:+list_for_each_entry_safe(page,next,list,lru)+__free_pages(page,0);+return-ENOMEM;+}++/**+*vmemmap_remap_alloc-remapthevmemmapvirtualaddressrange[@start,end)+*tothepagewhichisfromthe@vmemmap_pages+*respectively.+*@start:startaddressofthevmemmapvirtualaddressrangethatwewant+*toremap.+*@end:endaddressofthevmemmapvirtualaddressrangethatwewantto+*remap.+*@reuse:reuseaddress.+*@gpf_mask:GFPflagforallocatingvmemmappages.+*/+intvmemmap_remap_alloc(unsignedlongstart,unsignedlongend,+unsignedlongreuse,gfp_tgfp_mask)+{+LIST_HEAD(vmemmap_pages);+structvmemmap_remap_walkwalk={+.remap_pte=vmemmap_restore_pte,+.reuse_addr=reuse,+.vmemmap_pages=&vmemmap_pages,+};++/* See the comment in the vmemmap_remap_free(). */+BUG_ON(start-reuse!=PAGE_SIZE);++might_sleep_if(gfpflags_allow_blocking(gfp_mask));++if(alloc_vmemmap_page_list(start,end,gfp_mask,&vmemmap_pages))+return-ENOMEM;++vmemmap_remap_range(reuse,end,&walk);++return0;+}+/**Allocateablockofmemorytobeusedtobackthevirtualmemorymap*ortobackthepagetablesthatareusedtocreatethemapping.
Move bootmem info registration common API to individual bootmem_info.c.
And we will use {get,put}_page_bootmem() to initialize the page for the
vmemmap pages or free the vmemmap pages to buddy in the later patch.
So move them out of CONFIG_MEMORY_HOTPLUG_SPARSE. This is just code
movement without any functional change.
Signed-off-by: Muchun Song <redacted>
Acked-by: Mike Kravetz <redacted>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Reviewed-by: David Hildenbrand <redacted>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
From: Muchun Song <hidden> Date: 2021-03-03 05:38:40
On Wed, Mar 3, 2021 at 10:04 AM Mike Kravetz [off-list ref] wrote:
On 2/25/21 5:21 AM, Muchun Song wrote:
quoted
When we free a HugeTLB page to the buddy allocator, we should allocate
the vmemmap pages associated with it. But we may cannot allocate vmemmap
pages when the system is under memory pressure, in this case, we just
refuse to free the HugeTLB page instead of looping forever trying to
allocate the pages. This changes some behavior (list below) on some
cassociated with it.orner cases.
Suggest rewording this as:
When we free a HugeTLB page to the buddy allocator, we need to allocate
the vmemmap pages associated with it. However, we may not be able to
allocate the vmemmap pages when the system is under memory pressure. In
this case, we just refuse to free the HugeTLB page. This changes behavior
in some corner cases as listed below:
Thanks Mike. I will use this.
quoted
1) Failing to free a huge page triggered by the user (decrease nr_pages).
Need try again later by the user.
User needs to try again later.
quoted
2) Failing to free a surplus huge page when freed by the application.
Try again later when freeing a huge page next time.
3) Failing to dissolve a free huge page on ZONE_MOVABLE via
offline_pages().
This is a bit unfortunate if we have plenty of ZONE_MOVABLE memory
but are low on kernel memory. For example, migration of huge pages
would still work, however, dissolving the free page does not work.
This is a corner cases. When the system is that much under memory
pressure, offlining/unplug can be expected to fail. This is
unfortunate because it prevents from the memory offlining which
shouldn't happen for movable zones. People depending on the memory
hotplug and movable zone should carefuly consider whether savings
on unmovable memory are worth losing their hotplug functionality
in some situations.
Possible wording change:
This can happen when we have plenty of ZONE_MOVABLE memory, but
not enough kernel memory to allocate vmemmmap pages. We may even
be able to migrate huge page contents, but will not be able to
dissolve the source huge page. This will prevent an offline
operation and is unfortunate as memory offlining is expected to
succeed on movable zones. Users that depend on memory hotplug
to succeed for movable zones should carefully consider whether the
memory savings gained from this feature are worth the risk of
possibly not being able to offline memory in certain situations.
OK. Will use. Thanks.
quoted
4) Failing to dissolve a huge page on CMA/ZONE_MOVABLE via
alloc_contig_range() - once we have that handling in place. Mainly
affects CMA and virtio-mem.
Similar to 3). virito-mem will handle migration errors gracefully.
CMA might be able to fallback on other free areas within the CMA
region.
Vmemmap pages are allocated from the page freeing context. In order for
those allocations to be not disruptive (e.g. trigger oom killer)
__GFP_NORETRY is used. hugetlb_lock is dropped for the allocation
because a non sleeping allocation would be too fragile and it could fail
too easily under memory pressure. GFP_ATOMIC or other modes to access
memory reserves is not used because we want to prevent consuming
reserves under heavy hugetlb freeing.
Signed-off-by: Muchun Song <redacted>
---
Documentation/admin-guide/mm/hugetlbpage.rst | 8 +++
include/linux/mm.h | 2 +
mm/hugetlb.c | 92 +++++++++++++++++++++-------
mm/hugetlb_vmemmap.c | 32 ++++++----
mm/hugetlb_vmemmap.h | 23 +++++++
mm/sparse-vmemmap.c | 75 ++++++++++++++++++++++-
6 files changed, 197 insertions(+), 35 deletions(-)
@@ -60,6 +60,10 @@ HugePages_Surp the pool above the value in ``/proc/sys/vm/nr_hugepages``. The maximum number of surplus huge pages is controlled by``/proc/sys/vm/nr_overcommit_hugepages``.+ Note: When the feature of freeing unused vmemmap pages associated+ with each hugetlb page is enabled, the number of surplus huge pages+ may be temporarily larger than the maximum number of surplus huge+ pages when the system is under memory pressure. Hugepagesize is the default hugepage size (in Kb). Hugetlb
@@ -80,6 +84,10 @@ returned to the huge page pool when freed by a task. A user with root privileges can dynamically allocate more or free some persistent huge pages by increasing or decreasing the value of ``nr_hugepages``.+Note: When the feature of freeing unused vmemmap pages associated with each+hugetlb page is enabled, we can fail to free the huge pages triggered by+the user when ths system is under memory pressure. Please try again later.+ Pages that are used as huge pages are reserved inside the kernel and cannot be used for other purposes. Huge pages cannot be swapped out under memory pressure.
Thanks for your reminder. I will take a closer look at what the problem is.
This discussion/issue is independent of this patch. Since that issue
deals with existing code, we will need to come up with a solution there
first. A solution there may impact how free_huge_page is structured and
may impact this patch.
The in_atomic() check is insufficient to handle all cases. It is better
than !in_task(), but still does not cover all cases.
The rest of the patch looks good to me.
--
Mike Kravetz
quoted
/*
* Only call schedule_work() if hpage_freelist is previously
* empty. Otherwise, schedule_work() had been called but the
@@ -1699,8 +1738,7 @@ static int free_pool_huge_page(struct hstate *h, nodemask_t *nodes_allowed, h->surplus_huge_pages--; h->surplus_huge_pages_node[node]--; }- update_and_free_page(h, page);- ret = 1;+ ret = !update_and_free_page(h, page); break; } }
@@ -1713,10 +1751,14 @@ static int free_pool_huge_page(struct hstate *h, nodemask_t *nodes_allowed, * nothing for in-use hugepages and non-hugepages. * This function returns values like below: *- * -EBUSY: failed to dissolved free hugepages or the hugepage is in-use- * (allocated or reserved.)- * 0: successfully dissolved free hugepages or the page is not a- * hugepage (considered as already dissolved)+ * -ENOMEM: failed to allocate vmemmap pages to free the freed hugepages+ * when the system is under memory pressure and the feature of+ * freeing unused vmemmap pages associated with each hugetlb page+ * is enabled.+ * -EBUSY: failed to dissolved free hugepages or the hugepage is in-use+ * (allocated or reserved.)+ * 0: successfully dissolved free hugepages or the page is not a+ * hugepage (considered as already dissolved) */ int dissolve_free_huge_page(struct page *page) {
@@ -237,6 +238,78 @@ void vmemmap_remap_free(unsigned long start, unsigned long end,free_vmemmap_page_list(&vmemmap_pages);}+staticvoidvmemmap_restore_pte(pte_t*pte,unsignedlongaddr,+structvmemmap_remap_walk*walk)+{+pgprot_tpgprot=PAGE_KERNEL;+structpage*page;+void*to;++BUG_ON(pte_page(*pte)!=walk->reuse_page);++page=list_first_entry(walk->vmemmap_pages,structpage,lru);+list_del(&page->lru);+to=page_to_virt(page);+copy_page(to,(void*)walk->reuse_addr);++set_pte_at(&init_mm,addr,pte,mk_pte(page,pgprot));+}++staticintalloc_vmemmap_page_list(unsignedlongstart,unsignedlongend,+gfp_tgfp_mask,structlist_head*list)+{+unsignedlongnr_pages=(end-start)>>PAGE_SHIFT;+intnid=page_to_nid((structpage*)start);+structpage*page,*next;++while(nr_pages--){+page=alloc_pages_node(nid,gfp_mask,0);+if(!page)+gotoout;+list_add_tail(&page->lru,list);+}++return0;+out:+list_for_each_entry_safe(page,next,list,lru)+__free_pages(page,0);+return-ENOMEM;+}++/**+*vmemmap_remap_alloc-remapthevmemmapvirtualaddressrange[@start,end)+*tothepagewhichisfromthe@vmemmap_pages+*respectively.+*@start:startaddressofthevmemmapvirtualaddressrangethatwewant+*toremap.+*@end:endaddressofthevmemmapvirtualaddressrangethatwewantto+*remap.+*@reuse:reuseaddress.+*@gpf_mask:GFPflagforallocatingvmemmappages.+*/+intvmemmap_remap_alloc(unsignedlongstart,unsignedlongend,+unsignedlongreuse,gfp_tgfp_mask)+{+LIST_HEAD(vmemmap_pages);+structvmemmap_remap_walkwalk={+.remap_pte=vmemmap_restore_pte,+.reuse_addr=reuse,+.vmemmap_pages=&vmemmap_pages,+};++/* See the comment in the vmemmap_remap_free(). */+BUG_ON(start-reuse!=PAGE_SIZE);++might_sleep_if(gfpflags_allow_blocking(gfp_mask));++if(alloc_vmemmap_page_list(start,end,gfp_mask,&vmemmap_pages))+return-ENOMEM;++vmemmap_remap_range(reuse,end,&walk);++return0;+}+/**Allocateablockofmemorytobeusedtobackthevirtualmemorymap*ortobackthepagetablesthatareusedtocreatethemapping.
Hi all,
This patch series will free some vmemmap pages(struct page structures)
associated with each hugetlbpage when preallocated to save memory.
In order to reduce the difficulty of the first version of code review.
From this version, we disable PMD/huge page mapping of vmemmap if this
feature was enabled. This accutualy eliminate a bunch of the complex code
doing page table manipulation. When this patch series is solid, we cam add
the code of vmemmap page table manipulation in the future.
The struct page structures (page structs) are used to describe a physical
page frame. By default, there is a one-to-one mapping from a page frame to
it's corresponding page struct.
The HugeTLB pages consist of multiple base page size pages and is supported
by many architectures. See hugetlbpage.rst in the Documentation directory
for more details. On the x86 architecture, HugeTLB pages of size 2MB and 1GB
are currently supported. Since the base page size on x86 is 4KB, a 2MB
HugeTLB page consists of 512 base pages and a 1GB HugeTLB page consists of
4096 base pages. For each base page, there is a corresponding page struct.
Within the HugeTLB subsystem, only the first 4 page structs are used to
contain unique information about a HugeTLB page. HUGETLB_CGROUP_MIN_ORDER
provides this upper limit. The only 'useful' information in the remaining
page structs is the compound_head field, and this field is the same for all
tail pages.
The HUGETLB_CGROUP_MIN_ORDER is only when CGROUP_HUGETLB is enabled, but I guess
that does not matter
By removing redundant page structs for HugeTLB pages, memory can returned to
the buddy allocator for other uses.
When the system boot up, every 2M HugeTLB has 512 struct page structs which
size is 8 pages(sizeof(struct page) * 512 / PAGE_SIZE).
HugeTLB struct pages(8 pages) page frame(8 pages)
+-----------+ ---virt_to_page---> +-----------+ mapping to +-----------+
| | | 0 | -------------> | 0 |
| | +-----------+ +-----------+
| | | 1 | -------------> | 1 |
| | +-----------+ +-----------+
| | | 2 | -------------> | 2 |
| | +-----------+ +-----------+
| | | 3 | -------------> | 3 |
| | +-----------+ +-----------+
| | | 4 | -------------> | 4 |
| 2MB | +-----------+ +-----------+
| | | 5 | -------------> | 5 |
| | +-----------+ +-----------+
| | | 6 | -------------> | 6 |
| | +-----------+ +-----------+
| | | 7 | -------------> | 7 |
| | +-----------+ +-----------+
| |
| |
| |
+-----------+
The value of page->compound_head is the same for all tail pages. The first
page of page structs (page 0) associated with the HugeTLB page contains the 4
page structs necessary to describe the HugeTLB. The only use of the remaining
pages of page structs (page 1 to page 7) is to point to page->compound_head.
Therefore, we can remap pages 2 to 7 to page 1. Only 2 pages of page structs
will be used for each HugeTLB page. This will allow us to free the remaining
6 pages to the buddy allocator.
What is page 1 used for? page 0 carries the 4 struct pages needed, does compound_head
need a full page? IOW, why do we need two full pages -- may be the patches have the
answer to something I am missing?
Can these 6 pages come from the hugeTLB page itself? When you say 6 pages,
I presume you mean 6 pages of PAGE_SIZE
Apart from 2MB HugeTLB page, we also have 1GB HugeTLB page. It is similar
to the 2MB HugeTLB page. We also can use this approach to free the vmemmap
pages.
In this case, for the 1GB HugeTLB page, we can save 4094 pages. This is a
very substantial gain. On our server, run some SPDK/QEMU applications which
will use 1024GB hugetlbpage. With this feature enabled, we can save ~16GB
(1G hugepage)/~12GB (2MB hugepage) memory.
From: Muchun Song <hidden> Date: 2021-03-04 03:38:37
On Thu, Mar 4, 2021 at 11:14 AM Singh, Balbir [off-list ref] wrote:
On 26/2/21 12:21 am, Muchun Song wrote:
quoted
Hi all,
This patch series will free some vmemmap pages(struct page structures)
associated with each hugetlbpage when preallocated to save memory.
In order to reduce the difficulty of the first version of code review.
From this version, we disable PMD/huge page mapping of vmemmap if this
feature was enabled. This accutualy eliminate a bunch of the complex code
doing page table manipulation. When this patch series is solid, we cam add
the code of vmemmap page table manipulation in the future.
The struct page structures (page structs) are used to describe a physical
page frame. By default, there is a one-to-one mapping from a page frame to
it's corresponding page struct.
The HugeTLB pages consist of multiple base page size pages and is supported
by many architectures. See hugetlbpage.rst in the Documentation directory
for more details. On the x86 architecture, HugeTLB pages of size 2MB and 1GB
are currently supported. Since the base page size on x86 is 4KB, a 2MB
HugeTLB page consists of 512 base pages and a 1GB HugeTLB page consists of
4096 base pages. For each base page, there is a corresponding page struct.
Within the HugeTLB subsystem, only the first 4 page structs are used to
contain unique information about a HugeTLB page. HUGETLB_CGROUP_MIN_ORDER
provides this upper limit. The only 'useful' information in the remaining
page structs is the compound_head field, and this field is the same for all
tail pages.
The HUGETLB_CGROUP_MIN_ORDER is only when CGROUP_HUGETLB is enabled, but I guess
that does not matter
Agree.
quoted
By removing redundant page structs for HugeTLB pages, memory can returned to
the buddy allocator for other uses.
When the system boot up, every 2M HugeTLB has 512 struct page structs which
size is 8 pages(sizeof(struct page) * 512 / PAGE_SIZE).
HugeTLB struct pages(8 pages) page frame(8 pages)
+-----------+ ---virt_to_page---> +-----------+ mapping to +-----------+
| | | 0 | -------------> | 0 |
| | +-----------+ +-----------+
| | | 1 | -------------> | 1 |
| | +-----------+ +-----------+
| | | 2 | -------------> | 2 |
| | +-----------+ +-----------+
| | | 3 | -------------> | 3 |
| | +-----------+ +-----------+
| | | 4 | -------------> | 4 |
| 2MB | +-----------+ +-----------+
| | | 5 | -------------> | 5 |
| | +-----------+ +-----------+
| | | 6 | -------------> | 6 |
| | +-----------+ +-----------+
| | | 7 | -------------> | 7 |
| | +-----------+ +-----------+
| |
| |
| |
+-----------+
The value of page->compound_head is the same for all tail pages. The first
page of page structs (page 0) associated with the HugeTLB page contains the 4
page structs necessary to describe the HugeTLB. The only use of the remaining
pages of page structs (page 1 to page 7) is to point to page->compound_head.
Therefore, we can remap pages 2 to 7 to page 1. Only 2 pages of page structs
will be used for each HugeTLB page. This will allow us to free the remaining
6 pages to the buddy allocator.
What is page 1 used for? page 0 carries the 4 struct pages needed, does compound_head
need a full page? IOW, why do we need two full pages -- may be the patches have the
answer to something I am missing?
Yeah. It really can free 7 pages. But we need some work to support this. Why?
Now for the 2MB HugeTLB page, we only free 6 vmemmap pages. we really can
free 7 vmemmap pages. In this case, we can see 8 of the 512 struct page
structures have been set PG_head flag. If we can adjust compound_head()
slightly and make compound_head() return the real head struct page when
the parameter is the tail struct page but with PG_head flag set.
In order to make the code evolution route clearer. This feature can be
a separate patch (and send it out) after this patchset is solid and applied.
Apart from 2MB HugeTLB page, we also have 1GB HugeTLB page. It is similar
to the 2MB HugeTLB page. We also can use this approach to free the vmemmap
pages.
In this case, for the 1GB HugeTLB page, we can save 4094 pages. This is a
very substantial gain. On our server, run some SPDK/QEMU applications which
will use 1024GB hugetlbpage. With this feature enabled, we can save ~16GB
(1G hugepage)/~12GB (2MB hugepage) memory.
On Thu, Mar 04, 2021 at 11:36:44AM +0800, Muchun Song wrote:
On Thu, Mar 4, 2021 at 11:14 AM Singh, Balbir [off-list ref] wrote:
quoted
On 26/2/21 12:21 am, Muchun Song wrote:
quoted
Hi all,
This patch series will free some vmemmap pages(struct page structures)
associated with each hugetlbpage when preallocated to save memory.
In order to reduce the difficulty of the first version of code review.
From this version, we disable PMD/huge page mapping of vmemmap if this
feature was enabled. This accutualy eliminate a bunch of the complex code
doing page table manipulation. When this patch series is solid, we cam add
the code of vmemmap page table manipulation in the future.
The struct page structures (page structs) are used to describe a physical
page frame. By default, there is a one-to-one mapping from a page frame to
it's corresponding page struct.
The HugeTLB pages consist of multiple base page size pages and is supported
by many architectures. See hugetlbpage.rst in the Documentation directory
for more details. On the x86 architecture, HugeTLB pages of size 2MB and 1GB
are currently supported. Since the base page size on x86 is 4KB, a 2MB
HugeTLB page consists of 512 base pages and a 1GB HugeTLB page consists of
4096 base pages. For each base page, there is a corresponding page struct.
Within the HugeTLB subsystem, only the first 4 page structs are used to
contain unique information about a HugeTLB page. HUGETLB_CGROUP_MIN_ORDER
provides this upper limit. The only 'useful' information in the remaining
page structs is the compound_head field, and this field is the same for all
tail pages.
The HUGETLB_CGROUP_MIN_ORDER is only when CGROUP_HUGETLB is enabled, but I guess
that does not matter
Agree.
quoted
quoted
By removing redundant page structs for HugeTLB pages, memory can returned to
the buddy allocator for other uses.
When the system boot up, every 2M HugeTLB has 512 struct page structs which
size is 8 pages(sizeof(struct page) * 512 / PAGE_SIZE).
HugeTLB struct pages(8 pages) page frame(8 pages)
+-----------+ ---virt_to_page---> +-----------+ mapping to +-----------+
| | | 0 | -------------> | 0 |
| | +-----------+ +-----------+
| | | 1 | -------------> | 1 |
| | +-----------+ +-----------+
| | | 2 | -------------> | 2 |
| | +-----------+ +-----------+
| | | 3 | -------------> | 3 |
| | +-----------+ +-----------+
| | | 4 | -------------> | 4 |
| 2MB | +-----------+ +-----------+
| | | 5 | -------------> | 5 |
| | +-----------+ +-----------+
| | | 6 | -------------> | 6 |
| | +-----------+ +-----------+
| | | 7 | -------------> | 7 |
| | +-----------+ +-----------+
| |
| |
| |
+-----------+
The value of page->compound_head is the same for all tail pages. The first
page of page structs (page 0) associated with the HugeTLB page contains the 4
page structs necessary to describe the HugeTLB. The only use of the remaining
pages of page structs (page 1 to page 7) is to point to page->compound_head.
Therefore, we can remap pages 2 to 7 to page 1. Only 2 pages of page structs
will be used for each HugeTLB page. This will allow us to free the remaining
6 pages to the buddy allocator.
What is page 1 used for? page 0 carries the 4 struct pages needed, does compound_head
need a full page? IOW, why do we need two full pages -- may be the patches have the
answer to something I am missing?
Yeah. It really can free 7 pages. But we need some work to support this. Why?
Now for the 2MB HugeTLB page, we only free 6 vmemmap pages. we really can
free 7 vmemmap pages. In this case, we can see 8 of the 512 struct page
structures have been set PG_head flag. If we can adjust compound_head()
slightly and make compound_head() return the real head struct page when
the parameter is the tail struct page but with PG_head flag set.
In order to make the code evolution route clearer. This feature can be
a separate patch (and send it out) after this patchset is solid and applied.
On Wed, Mar 03, 2021 at 01:45:00PM +1100, Singh, Balbir wrote:
On 26/2/21 12:21 am, Muchun Song wrote:
quoted
Move bootmem info registration common API to individual bootmem_info.c.
And we will use {get,put}_page_bootmem() to initialize the page for the
vmemmap pages or free the vmemmap pages to buddy in the later patch.
So move them out of CONFIG_MEMORY_HOTPLUG_SPARSE. This is just code
movement without any functional change.
Signed-off-by: Muchun Song <redacted>
Acked-by: Mike Kravetz <redacted>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Reviewed-by: David Hildenbrand <redacted>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
On Thu, Feb 25, 2021 at 09:21:23PM +0800, Muchun Song wrote:
The option HUGETLB_PAGE_FREE_VMEMMAP allows for the freeing of
some vmemmap pages associated with pre-allocated HugeTLB pages.
For example, on X86_64 6 vmemmap pages of size 4KB each can be
saved for each 2MB HugeTLB page. 4094 vmemmap pages of size 4KB
each can be saved for each 1GB HugeTLB page.
When a HugeTLB page is allocated or freed, the vmemmap array
representing the range associated with the page will need to be
remapped. When a page is allocated, vmemmap pages are freed
after remapping. When a page is freed, previously discarded
vmemmap pages must be allocated before remapping.
The config option is introduced early so that supporting code
can be written to depend on the option. The initial version of
the code only provides support for x86-64.
Like other code which frees vmemmap, this config option depends on
HAVE_BOOTMEM_INFO_NODE. The routine register_page_bootmem_info() is
used to register bootmem info. Therefore, make sure
register_page_bootmem_info is enabled if HUGETLB_PAGE_FREE_VMEMMAP
is defined.
Signed-off-by: Muchun Song <redacted>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Acked-by: Mike Kravetz <redacted>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
---
From: Muchun Song <hidden> Date: 2021-03-04 05:52:41
On Thu, Mar 4, 2021 at 12:26 PM Balbir Singh [off-list ref] wrote:
On Wed, Mar 03, 2021 at 01:45:00PM +1100, Singh, Balbir wrote:
quoted
On 26/2/21 12:21 am, Muchun Song wrote:
quoted
Move bootmem info registration common API to individual bootmem_info.c.
And we will use {get,put}_page_bootmem() to initialize the page for the
vmemmap pages or free the vmemmap pages to buddy in the later patch.
So move them out of CONFIG_MEMORY_HOTPLUG_SPARSE. This is just code
movement without any functional change.
Signed-off-by: Muchun Song <redacted>
Acked-by: Mike Kravetz <redacted>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Reviewed-by: David Hildenbrand <redacted>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
Every HugeTLB has more than one struct page structure. We __know__ that
we only use the first 4(HUGETLB_CGROUP_MIN_ORDER) struct page structures
to store metadata associated with each HugeTLB.
There are a lot of struct page structures associated with each HugeTLB
page. For tail pages, the value of compound_head is the same. So we can
reuse first page of tail page structures. We map the virtual addresses
of the remaining pages of tail page structures to the first tail page
struct, and then free these page frames. Therefore, we need to reserve
two pages as vmemmap areas.
When we allocate a HugeTLB page from the buddy, we can free some vmemmap
pages associated with each HugeTLB page. It is more appropriate to do it
in the prep_new_huge_page().
The free_vmemmap_pages_per_hpage(), which indicates how many vmemmap
pages associated with a HugeTLB page can be freed, returns zero for
now, which means the feature is disabled. We will enable it once all
the infrastructure is there.
Signed-off-by: Muchun Song <redacted>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
---
include/linux/bootmem_info.h | 27 +++++-
include/linux/mm.h | 3 +
mm/Makefile | 1 +
mm/hugetlb.c | 3 +
mm/hugetlb_vmemmap.c | 219 +++++++++++++++++++++++++++++++++++++++++++
mm/hugetlb_vmemmap.h | 20 ++++
mm/sparse-vmemmap.c | 207 ++++++++++++++++++++++++++++++++++++++++
7 files changed, 479 insertions(+), 1 deletion(-)
create mode 100644 mm/hugetlb_vmemmap.c
create mode 100644 mm/hugetlb_vmemmap.h
The concurrency semantics of this code are not clear, do we need READ_ONCE()/
WRITE_ONCE() semantics if this page walk is lockless? Can we run this code
in parallel on the same section? I presume not
+ /*
+ * Because the reuse address is part of the range that we are
+ * walking, skip the reuse address range.
+ */
+ addr += PAGE_SIZE;
+ }
+
+ for (; addr != end; addr += PAGE_SIZE, pte++) {
+ BUG_ON(pte_none(*pte));
+
+ walk->remap_pte(pte, addr, walk);
+ }
+}
+
+static void vmemmap_pmd_range(pud_t *pud, unsigned long addr,
+ unsigned long end,
+ struct vmemmap_remap_walk *walk)
+{
+ pmd_t *pmd;
+ unsigned long next;
+
+ pmd = pmd_offset(pud, addr);
+ do {
+ BUG_ON(pmd_none(*pmd) || pmd_leaf(*pmd));
+
+ next = pmd_addr_end(addr, end);
+ vmemmap_pte_range(pmd, addr, next, walk);
+ } while (pmd++, addr = next, addr != end);
+}
+
+static void vmemmap_pud_range(p4d_t *p4d, unsigned long addr,
+ unsigned long end,
+ struct vmemmap_remap_walk *walk)
+{
+ pud_t *pud;
+ unsigned long next;
+
+ pud = pud_offset(p4d, addr);
+ do {
+ BUG_ON(pud_none(*pud));
+
+ next = pud_addr_end(addr, end);
+ vmemmap_pmd_range(pud, addr, next, walk);
+ } while (pud++, addr = next, addr != end);
+}
+
+static void vmemmap_p4d_range(pgd_t *pgd, unsigned long addr,
+ unsigned long end,
+ struct vmemmap_remap_walk *walk)
+{
+ p4d_t *p4d;
+ unsigned long next;
+
+ p4d = p4d_offset(pgd, addr);
+ do {
+ BUG_ON(p4d_none(*p4d));
+
+ next = p4d_addr_end(addr, end);
+ vmemmap_pud_range(p4d, addr, next, walk);
+ } while (p4d++, addr = next, addr != end);
+}
+
+static void vmemmap_remap_range(unsigned long start, unsigned long end,
+ struct vmemmap_remap_walk *walk)
+{
+ unsigned long addr = start;
+ unsigned long next;
+ pgd_t *pgd;
+
+ VM_BUG_ON(!IS_ALIGNED(start, PAGE_SIZE));
+ VM_BUG_ON(!IS_ALIGNED(end, PAGE_SIZE));
+
+ pgd = pgd_offset_k(addr);
+ do {
+ BUG_ON(pgd_none(*pgd));
+
+ next = pgd_addr_end(addr, end);
+ vmemmap_p4d_range(pgd, addr, next, walk);
+ } while (pgd++, addr = next, addr != end);
+
+ /*
+ * We only change the mapping of the vmemmap virtual address range
+ * [@start + PAGE_SIZE, end), so we only need to flush the TLB which
+ * belongs to the range.
+ */
+ flush_tlb_kernel_range(start + PAGE_SIZE, end);
+}
+
+/*
+ * Free a vmemmap page. A vmemmap page can be allocated from the memblock
+ * allocator or buddy allocator. If the PG_reserved flag is set, it means
+ * that it allocated from the memblock allocator, just free it via the
+ * free_bootmem_page(). Otherwise, use __free_page().
+ */
+static inline void free_vmemmap_page(struct page *page)
+{
+ if (PageReserved(page))
+ free_bootmem_page(page);
+ else
+ __free_page(page);
+}
+
+/* Free a list of the vmemmap pages */
+static void free_vmemmap_page_list(struct list_head *list)
+{
+ struct page *page, *next;
+
+ list_for_each_entry_safe(page, next, list, lru) {
+ list_del(&page->lru);
+ free_vmemmap_page(page);
+ }
+}
+
+static void vmemmap_remap_pte(pte_t *pte, unsigned long addr,
+ struct vmemmap_remap_walk *walk)
+{
+ /*
+ * Remap the tail pages as read-only to catch illegal write operation
+ * to the tail pages.
+ */
+ pgprot_t pgprot = PAGE_KERNEL_RO;
+ pte_t entry = mk_pte(walk->reuse_page, pgprot);
+ struct page *page = pte_page(*pte);
+
+ list_add(&page->lru, walk->vmemmap_pages);
+ set_pte_at(&init_mm, addr, pte, entry);
+}
+
+/**
+ * vmemmap_remap_free - remap the vmemmap virtual address range [@start, @end)
+ * to the page which @reuse is mapped to, then free vmemmap
+ * which the range are mapped to.
+ * @start: start address of the vmemmap virtual address range that we want
+ * to remap.
+ * @end: end address of the vmemmap virtual address range that we want to
+ * remap.
+ * @reuse: reuse address.
+ *
+ * Note: This function depends on vmemmap being base page mapped. Please make
+ * sure that we disable PMD mapping of vmemmap pages when calling this function.
This is something that the walking code enforces via BUG_ON's right?
+ */
+void vmemmap_remap_free(unsigned long start, unsigned long end,
+ unsigned long reuse)
+{
+ LIST_HEAD(vmemmap_pages);
+ struct vmemmap_remap_walk walk = {
+ .remap_pte = vmemmap_remap_pte,
+ .reuse_addr = reuse,
+ .vmemmap_pages = &vmemmap_pages,
+ };
+
+ /*
+ * In order to make remapping routine most efficient for the huge pages,
+ * the routine of vmemmap page table walking has the following rules
+ * (see more details from the vmemmap_pte_range()):
+ *
+ * - The range [@start, @end) and the range [@reuse, @reuse + PAGE_SIZE)
+ * should be continuous.
+ * - The @reuse address is part of the range [@reuse, @end) that we are
+ * walking which is passed to vmemmap_remap_range().
+ * - The @reuse address is the first in the complete range.
+ *
+ * So we need to make sure that @start and @reuse meet the above rules.
+ */
+ BUG_ON(start - reuse != PAGE_SIZE);
Why even take a reuse arg then, just set reuse = start - PAGE_SIZE? If we do that
we can rename the function to reflect that the second page is reused or keep this
function and create an inline wrapper with reuse set to start - PAGE_SIZE and use
that for this use case and remove this BUG_ON
+
+ vmemmap_remap_range(reuse, end, &walk);
+ free_vmemmap_page_list(&vmemmap_pages);
+}
/*
* Allocate a block of memory to be used to back the virtual memory map
From: Muchun Song <hidden> Date: 2021-03-05 04:42:10
On Fri, Mar 5, 2021 at 7:50 AM Singh, Balbir [off-list ref] wrote:
On 26/2/21 12:21 am, Muchun Song wrote:
quoted
Every HugeTLB has more than one struct page structure. We __know__ that
we only use the first 4(HUGETLB_CGROUP_MIN_ORDER) struct page structures
to store metadata associated with each HugeTLB.
There are a lot of struct page structures associated with each HugeTLB
page. For tail pages, the value of compound_head is the same. So we can
reuse first page of tail page structures. We map the virtual addresses
of the remaining pages of tail page structures to the first tail page
struct, and then free these page frames. Therefore, we need to reserve
two pages as vmemmap areas.
When we allocate a HugeTLB page from the buddy, we can free some vmemmap
pages associated with each HugeTLB page. It is more appropriate to do it
in the prep_new_huge_page().
The free_vmemmap_pages_per_hpage(), which indicates how many vmemmap
pages associated with a HugeTLB page can be freed, returns zero for
now, which means the feature is disabled. We will enable it once all
the infrastructure is there.
Signed-off-by: Muchun Song <redacted>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
---
include/linux/bootmem_info.h | 27 +++++-
include/linux/mm.h | 3 +
mm/Makefile | 1 +
mm/hugetlb.c | 3 +
mm/hugetlb_vmemmap.c | 219 +++++++++++++++++++++++++++++++++++++++++++
mm/hugetlb_vmemmap.h | 20 ++++
mm/sparse-vmemmap.c | 207 ++++++++++++++++++++++++++++++++++++++++
7 files changed, 479 insertions(+), 1 deletion(-)
create mode 100644 mm/hugetlb_vmemmap.c
create mode 100644 mm/hugetlb_vmemmap.h
The concurrency semantics of this code are not clear, do we need READ_ONCE()/
WRITE_ONCE() semantics if this page walk is lockless? Can we run this code
in parallel on the same section? I presume not
IIUC, there is no parallel thread to walk the page tables of the
vmemmap area. We may not need READ_ONCE/WRITE_ONCE.
quoted
+ /*
+ * Because the reuse address is part of the range that we are
+ * walking, skip the reuse address range.
+ */
+ addr += PAGE_SIZE;
+ }
+
+ for (; addr != end; addr += PAGE_SIZE, pte++) {
+ BUG_ON(pte_none(*pte));
+
+ walk->remap_pte(pte, addr, walk);
+ }
+}
+
+static void vmemmap_pmd_range(pud_t *pud, unsigned long addr,
+ unsigned long end,
+ struct vmemmap_remap_walk *walk)
+{
+ pmd_t *pmd;
+ unsigned long next;
+
+ pmd = pmd_offset(pud, addr);
+ do {
+ BUG_ON(pmd_none(*pmd) || pmd_leaf(*pmd));
+
+ next = pmd_addr_end(addr, end);
+ vmemmap_pte_range(pmd, addr, next, walk);
+ } while (pmd++, addr = next, addr != end);
+}
+
+static void vmemmap_pud_range(p4d_t *p4d, unsigned long addr,
+ unsigned long end,
+ struct vmemmap_remap_walk *walk)
+{
+ pud_t *pud;
+ unsigned long next;
+
+ pud = pud_offset(p4d, addr);
+ do {
+ BUG_ON(pud_none(*pud));
+
+ next = pud_addr_end(addr, end);
+ vmemmap_pmd_range(pud, addr, next, walk);
+ } while (pud++, addr = next, addr != end);
+}
+
+static void vmemmap_p4d_range(pgd_t *pgd, unsigned long addr,
+ unsigned long end,
+ struct vmemmap_remap_walk *walk)
+{
+ p4d_t *p4d;
+ unsigned long next;
+
+ p4d = p4d_offset(pgd, addr);
+ do {
+ BUG_ON(p4d_none(*p4d));
+
+ next = p4d_addr_end(addr, end);
+ vmemmap_pud_range(p4d, addr, next, walk);
+ } while (p4d++, addr = next, addr != end);
+}
+
+static void vmemmap_remap_range(unsigned long start, unsigned long end,
+ struct vmemmap_remap_walk *walk)
+{
+ unsigned long addr = start;
+ unsigned long next;
+ pgd_t *pgd;
+
+ VM_BUG_ON(!IS_ALIGNED(start, PAGE_SIZE));
+ VM_BUG_ON(!IS_ALIGNED(end, PAGE_SIZE));
+
+ pgd = pgd_offset_k(addr);
+ do {
+ BUG_ON(pgd_none(*pgd));
+
+ next = pgd_addr_end(addr, end);
+ vmemmap_p4d_range(pgd, addr, next, walk);
+ } while (pgd++, addr = next, addr != end);
+
+ /*
+ * We only change the mapping of the vmemmap virtual address range
+ * [@start + PAGE_SIZE, end), so we only need to flush the TLB which
+ * belongs to the range.
+ */
+ flush_tlb_kernel_range(start + PAGE_SIZE, end);
+}
+
+/*
+ * Free a vmemmap page. A vmemmap page can be allocated from the memblock
+ * allocator or buddy allocator. If the PG_reserved flag is set, it means
+ * that it allocated from the memblock allocator, just free it via the
+ * free_bootmem_page(). Otherwise, use __free_page().
+ */
+static inline void free_vmemmap_page(struct page *page)
+{
+ if (PageReserved(page))
+ free_bootmem_page(page);
+ else
+ __free_page(page);
+}
+
+/* Free a list of the vmemmap pages */
+static void free_vmemmap_page_list(struct list_head *list)
+{
+ struct page *page, *next;
+
+ list_for_each_entry_safe(page, next, list, lru) {
+ list_del(&page->lru);
+ free_vmemmap_page(page);
+ }
+}
+
+static void vmemmap_remap_pte(pte_t *pte, unsigned long addr,
+ struct vmemmap_remap_walk *walk)
+{
+ /*
+ * Remap the tail pages as read-only to catch illegal write operation
+ * to the tail pages.
+ */
+ pgprot_t pgprot = PAGE_KERNEL_RO;
+ pte_t entry = mk_pte(walk->reuse_page, pgprot);
+ struct page *page = pte_page(*pte);
+
+ list_add(&page->lru, walk->vmemmap_pages);
+ set_pte_at(&init_mm, addr, pte, entry);
+}
+
+/**
+ * vmemmap_remap_free - remap the vmemmap virtual address range [@start, @end)
+ * to the page which @reuse is mapped to, then free vmemmap
+ * which the range are mapped to.
+ * @start: start address of the vmemmap virtual address range that we want
+ * to remap.
+ * @end: end address of the vmemmap virtual address range that we want to
+ * remap.
+ * @reuse: reuse address.
+ *
+ * Note: This function depends on vmemmap being base page mapped. Please make
+ * sure that we disable PMD mapping of vmemmap pages when calling this function.
This is something that the walking code enforces via BUG_ON's right?
Right. There is a BUG_ON(pmd_leaf(*pmd)) in vmemmap_pmd_range().
quoted
+ */
+void vmemmap_remap_free(unsigned long start, unsigned long end,
+ unsigned long reuse)
+{
+ LIST_HEAD(vmemmap_pages);
+ struct vmemmap_remap_walk walk = {
+ .remap_pte = vmemmap_remap_pte,
+ .reuse_addr = reuse,
+ .vmemmap_pages = &vmemmap_pages,
+ };
+
+ /*
+ * In order to make remapping routine most efficient for the huge pages,
+ * the routine of vmemmap page table walking has the following rules
+ * (see more details from the vmemmap_pte_range()):
+ *
+ * - The range [@start, @end) and the range [@reuse, @reuse + PAGE_SIZE)
+ * should be continuous.
+ * - The @reuse address is part of the range [@reuse, @end) that we are
+ * walking which is passed to vmemmap_remap_range().
+ * - The @reuse address is the first in the complete range.
+ *
+ * So we need to make sure that @start and @reuse meet the above rules.
+ */
+ BUG_ON(start - reuse != PAGE_SIZE);
Why even take a reuse arg then, just set reuse = start - PAGE_SIZE? If we do that
we can rename the function to reflect that the second page is reused or
keep this
function and create an inline wrapper with reuse set to start - PAGE_SIZE and use
that for this use case and remove this BUG_ON
I also want to hear Oscar and Mike's suggestions about this.
Thanks.
quoted
+
+ vmemmap_remap_range(reuse, end, &walk);
+ free_vmemmap_page_list(&vmemmap_pages);
+}
/*
* Allocate a block of memory to be used to back the virtual memory map
On Thu, Feb 25, 2021 at 09:21:25PM +0800, Muchun Song wrote:
quoted hunk
When we free a HugeTLB page to the buddy allocator, we should allocate
the vmemmap pages associated with it. But we may cannot allocate vmemmap
pages when the system is under memory pressure, in this case, we just
refuse to free the HugeTLB page instead of looping forever trying to
allocate the pages. This changes some behavior (list below) on some
corner cases.
1) Failing to free a huge page triggered by the user (decrease nr_pages).
Need try again later by the user.
2) Failing to free a surplus huge page when freed by the application.
Try again later when freeing a huge page next time.
3) Failing to dissolve a free huge page on ZONE_MOVABLE via
offline_pages().
This is a bit unfortunate if we have plenty of ZONE_MOVABLE memory
but are low on kernel memory. For example, migration of huge pages
would still work, however, dissolving the free page does not work.
This is a corner cases. When the system is that much under memory
pressure, offlining/unplug can be expected to fail. This is
unfortunate because it prevents from the memory offlining which
shouldn't happen for movable zones. People depending on the memory
hotplug and movable zone should carefuly consider whether savings
on unmovable memory are worth losing their hotplug functionality
in some situations.
4) Failing to dissolve a huge page on CMA/ZONE_MOVABLE via
alloc_contig_range() - once we have that handling in place. Mainly
affects CMA and virtio-mem.
Similar to 3). virito-mem will handle migration errors gracefully.
CMA might be able to fallback on other free areas within the CMA
region.
Vmemmap pages are allocated from the page freeing context. In order for
those allocations to be not disruptive (e.g. trigger oom killer)
__GFP_NORETRY is used. hugetlb_lock is dropped for the allocation
because a non sleeping allocation would be too fragile and it could fail
too easily under memory pressure. GFP_ATOMIC or other modes to access
memory reserves is not used because we want to prevent consuming
reserves under heavy hugetlb freeing.
Signed-off-by: Muchun Song <redacted>
---
Documentation/admin-guide/mm/hugetlbpage.rst | 8 +++
include/linux/mm.h | 2 +
mm/hugetlb.c | 92 +++++++++++++++++++++-------
mm/hugetlb_vmemmap.c | 32 ++++++----
mm/hugetlb_vmemmap.h | 23 +++++++
mm/sparse-vmemmap.c | 75 ++++++++++++++++++++++-
6 files changed, 197 insertions(+), 35 deletions(-)
@@ -60,6 +60,10 @@ HugePages_Surp the pool above the value in ``/proc/sys/vm/nr_hugepages``. The maximum number of surplus huge pages is controlled by``/proc/sys/vm/nr_overcommit_hugepages``.+ Note: When the feature of freeing unused vmemmap pages associated+ with each hugetlb page is enabled, the number of surplus huge pages+ may be temporarily larger than the maximum number of surplus huge+ pages when the system is under memory pressure. Hugepagesize is the default hugepage size (in Kb). Hugetlb
@@ -80,6 +84,10 @@ returned to the huge page pool when freed by a task. A user with root privileges can dynamically allocate more or free some persistent huge pages by increasing or decreasing the value of ``nr_hugepages``.+Note: When the feature of freeing unused vmemmap pages associated with each+hugetlb page is enabled, we can fail to free the huge pages triggered by+the user when ths system is under memory pressure. Please try again later.+ Pages that are used as huge pages are reserved inside the kernel and cannot be used for other purposes. Huge pages cannot be swapped out under memory pressure.
@@ -1404,9 +1443,9 @@ static void __free_huge_page(struct page *page)}elseif(h->surplus_huge_pages_node[nid]){/* remove the page from active list */list_del(&page->lru);-update_and_free_page(h,page);h->surplus_huge_pages--;h->surplus_huge_pages_node[nid]--;+update_and_free_page(h,page);}else{arch_clear_hugepage_flags(page);enqueue_huge_page(h,page);
This is where I think some optimization is possible, once we are done with
vmemmap_end calculation, we can use 6 pages (for 2MiB huge page) as pages
for struct page. Is there a reason to not do so?
Balbir
From: Muchun Song <hidden> Date: 2021-03-05 09:32:03
On Fri, Mar 5, 2021 at 4:55 PM Balbir Singh [off-list ref] wrote:
On Thu, Feb 25, 2021 at 09:21:25PM +0800, Muchun Song wrote:
quoted
When we free a HugeTLB page to the buddy allocator, we should allocate
the vmemmap pages associated with it. But we may cannot allocate vmemmap
pages when the system is under memory pressure, in this case, we just
refuse to free the HugeTLB page instead of looping forever trying to
allocate the pages. This changes some behavior (list below) on some
corner cases.
1) Failing to free a huge page triggered by the user (decrease nr_pages).
Need try again later by the user.
2) Failing to free a surplus huge page when freed by the application.
Try again later when freeing a huge page next time.
3) Failing to dissolve a free huge page on ZONE_MOVABLE via
offline_pages().
This is a bit unfortunate if we have plenty of ZONE_MOVABLE memory
but are low on kernel memory. For example, migration of huge pages
would still work, however, dissolving the free page does not work.
This is a corner cases. When the system is that much under memory
pressure, offlining/unplug can be expected to fail. This is
unfortunate because it prevents from the memory offlining which
shouldn't happen for movable zones. People depending on the memory
hotplug and movable zone should carefuly consider whether savings
on unmovable memory are worth losing their hotplug functionality
in some situations.
4) Failing to dissolve a huge page on CMA/ZONE_MOVABLE via
alloc_contig_range() - once we have that handling in place. Mainly
affects CMA and virtio-mem.
Similar to 3). virito-mem will handle migration errors gracefully.
CMA might be able to fallback on other free areas within the CMA
region.
Vmemmap pages are allocated from the page freeing context. In order for
those allocations to be not disruptive (e.g. trigger oom killer)
__GFP_NORETRY is used. hugetlb_lock is dropped for the allocation
because a non sleeping allocation would be too fragile and it could fail
too easily under memory pressure. GFP_ATOMIC or other modes to access
memory reserves is not used because we want to prevent consuming
reserves under heavy hugetlb freeing.
Signed-off-by: Muchun Song <redacted>
---
Documentation/admin-guide/mm/hugetlbpage.rst | 8 +++
include/linux/mm.h | 2 +
mm/hugetlb.c | 92 +++++++++++++++++++++-------
mm/hugetlb_vmemmap.c | 32 ++++++----
mm/hugetlb_vmemmap.h | 23 +++++++
mm/sparse-vmemmap.c | 75 ++++++++++++++++++++++-
6 files changed, 197 insertions(+), 35 deletions(-)
@@ -60,6 +60,10 @@ HugePages_Surp the pool above the value in ``/proc/sys/vm/nr_hugepages``. The maximum number of surplus huge pages is controlled by``/proc/sys/vm/nr_overcommit_hugepages``.+ Note: When the feature of freeing unused vmemmap pages associated+ with each hugetlb page is enabled, the number of surplus huge pages+ may be temporarily larger than the maximum number of surplus huge+ pages when the system is under memory pressure. Hugepagesize is the default hugepage size (in Kb). Hugetlb
@@ -80,6 +84,10 @@ returned to the huge page pool when freed by a task. A user with root privileges can dynamically allocate more or free some persistent huge pages by increasing or decreasing the value of ``nr_hugepages``.+Note: When the feature of freeing unused vmemmap pages associated with each+hugetlb page is enabled, we can fail to free the huge pages triggered by+the user when ths system is under memory pressure. Please try again later.+ Pages that are used as huge pages are reserved inside the kernel and cannot be used for other purposes. Huge pages cannot be swapped out under memory pressure.
@@ -1404,9 +1443,9 @@ static void __free_huge_page(struct page *page)}elseif(h->surplus_huge_pages_node[nid]){/* remove the page from active list */list_del(&page->lru);-update_and_free_page(h,page);h->surplus_huge_pages--;h->surplus_huge_pages_node[nid]--;+update_and_free_page(h,page);}else{arch_clear_hugepage_flags(page);enqueue_huge_page(h,page);
This is where I think some optimization is possible, once we are done with
vmemmap_end calculation, we can use 6 pages (for 2MiB huge page) as pages
for struct page. Is there a reason to not do so?
On Thu, Feb 25, 2021 at 09:21:26PM +0800, Muchun Song wrote:
Because we reuse the first tail vmemmap page frame and remap it
with read-only, we cannot set the PageHWPosion on some tail pages.
So we can use the head[4].private (There are at least 128 struct
page structures associated with the optimized HugeTLB page, so
using head[4].private is safe) to record the real error page index
and set the raw error page PageHWPoison later.
Does the hardcoding of 4 come from HUGETLB_CGROUP_MIN_ORDER, if so
do we need to hardcode 4? Also, I am not sure about the comment
on safety and 128 struct pages
Balbir
From: Muchun Song <hidden> Date: 2021-03-07 08:40:32
On Sun, Mar 7, 2021 at 4:19 PM Balbir Singh [off-list ref] wrote:
On Thu, Feb 25, 2021 at 09:21:26PM +0800, Muchun Song wrote:
quoted
Because we reuse the first tail vmemmap page frame and remap it
with read-only, we cannot set the PageHWPosion on some tail pages.
So we can use the head[4].private (There are at least 128 struct
page structures associated with the optimized HugeTLB page, so
using head[4].private is safe) to record the real error page index
and set the raw error page PageHWPoison later.
Does the hardcoding of 4 come from HUGETLB_CGROUP_MIN_ORDER, if so
Yes.
do we need to hardcode 4? Also, I am not sure about the comment
on safety and 128 struct pages
We can set head[4].private only if free_vmemmap_pages_per_hpage(h)
returns true. In this case, there are 128 struct page structures (we reserve
2 pages as vmemmap pages, so 2 * 4KB / sizeof(struct page) == 128) that
can be used. Instead of hardcode, I introduce another patch to make the
code more readable. Please refer to patch #8 in this series.
Thanks.